Redis/redis-For-Windows
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 3 Wiki Activity

Compare commits

base: Redis/redis-For-Windows:3.0.1
Branches Tags
Redis/redis-For-Windows:develop Redis/redis-For-Windows:win-5.0 Redis/redis-For-Windows:win-6.2
Redis/redis-For-Windows:5.0.14.1 Redis/redis-For-Windows:v5.0.14.1 Redis/redis-For-Windows:v5.0.14 Redis/redis-For-Windows:v5.0.10 Redis/redis-For-Windows:v5.0.9 Redis/redis-For-Windows:v4.0.14.2 Redis/redis-For-Windows:v4.0.14.1 Redis/redis-For-Windows:v4.0.14 Redis/redis-For-Windows:v4.0.2.3-alpha Redis/redis-For-Windows:v4.0.2.2-alpha Redis/redis-For-Windows:v4.0.2.1-alpha Redis/redis-For-Windows:v4.0.2-alpha Redis/redis-For-Windows:win-3.2.100 Redis/redis-For-Windows:win-3.0.504 Redis/redis-For-Windows:win-3.0.503 Redis/redis-For-Windows:win-2.8.2402 Redis/redis-For-Windows:win-3.0.502 Redis/redis-For-Windows:win-2.8.2401 Redis/redis-For-Windows:3.2.1 Redis/redis-For-Windows:win-3.2.0000-preview Redis/redis-For-Windows:3.2.0 Redis/redis-For-Windows:3.2.0-rc3 Redis/redis-For-Windows:3.2.0-rc2 Redis/redis-For-Windows:win-2.8.2400 Redis/redis-For-Windows:win-3.0.501 Redis/redis-For-Windows:3.2-rc1 Redis/redis-For-Windows:3.0.6 Redis/redis-For-Windows:2.8.24 Redis/redis-For-Windows:win-3.0.500 Redis/redis-For-Windows:win-3.0.500-rc2 Redis/redis-For-Windows:win-3.0.500-rc1 Redis/redis-For-Windows:win-2.8.2104 Redis/redis-For-Windows:3.0.5 Redis/redis-For-Windows:2.8.23 Redis/redis-For-Windows:win-3.0.300-beta1 Redis/redis-For-Windows:win-2.8.2103 Redis/redis-For-Windows:2.8.22 Redis/redis-For-Windows:3.0.4 Redis/redis-For-Windows:win-3.0.300-alpha3 Redis/redis-For-Windows:win-2.8.2102 Redis/redis-For-Windows:win-3.0.300-alpha2 Redis/redis-For-Windows:win-3.0.100-alpha1 Redis/redis-For-Windows:3.0.3 Redis/redis-For-Windows:win-2.8.2101 Redis/redis-For-Windows:win-2.8.21 Redis/redis-For-Windows:3.0.2 Redis/redis-For-Windows:2.8.21 Redis/redis-For-Windows:2.8.20 Redis/redis-For-Windows:3.0.1 Redis/redis-For-Windows:win-2.8.19.1 Redis/redis-For-Windows:3.0.0 Redis/redis-For-Windows:3.0.0-rc6 Redis/redis-For-Windows:3.0.0-rc5 Redis/redis-For-Windows:win-2.8.19-rc1 Redis/redis-For-Windows:win-2.8.19 Redis/redis-For-Windows:3.0.0-rc4 Redis/redis-For-Windows:win-2.8.19-pre2 Redis/redis-For-Windows:win-2.8.17.4 Redis/redis-For-Windows:3.0.0-rc3 Redis/redis-For-Windows:3.0.0-rc2 Redis/redis-For-Windows:win-2.8.17.3 Redis/redis-For-Windows:win-2.8.19-pre1 Redis/redis-For-Windows:win-2.8.17.2 Redis/redis-For-Windows:win-2.8.17.1 Redis/redis-For-Windows:2.8.19 Redis/redis-For-Windows:2.8.18 Redis/redis-For-Windows:3.0.0-rc1 Redis/redis-For-Windows:2.8.17 Redis/redis-For-Windows:2.8.16 Redis/redis-For-Windows:2.8.15 Redis/redis-For-Windows:win2.8.12 Redis/redis-For-Windows:2.8.14 Redis/redis-For-Windows:3.0.0-beta8 Redis/redis-For-Windows:2.8.13 Redis/redis-For-Windows:3.0.0-beta7 Redis/redis-For-Windows:win_2.8.9 Redis/redis-For-Windows:2.8.12 Redis/redis-For-Windows:2.8.11 Redis/redis-For-Windows:3.0.0-beta6 Redis/redis-For-Windows:2.8.10 Redis/redis-For-Windows:3.0.0-beta5 Redis/redis-For-Windows:3.0.0-beta4 Redis/redis-For-Windows:2.6.14_msopentech Redis/redis-For-Windows:2.8.4_msopentech Redis/redis-For-Windows:msopentech_2.8.4 Redis/redis-For-Windows:3.0.0-beta3 Redis/redis-For-Windows:2.8.9 Redis/redis-For-Windows:2.8.8 Redis/redis-For-Windows:3.0.0-beta2 Redis/redis-For-Windows:2.8.7 Redis/redis-For-Windows:2.8.6 Redis/redis-For-Windows:3.0.0-beta1 Redis/redis-For-Windows:2.8.5 Redis/redis-For-Windows:2.8.4 Redis/redis-For-Windows:2.8.3 Redis/redis-For-Windows:2.8.2 Redis/redis-For-Windows:2.8.1 Redis/redis-For-Windows:2.8.0 Redis/redis-For-Windows:2.8.0-rc6 Redis/redis-For-Windows:2.8.0-rc5 Redis/redis-For-Windows:2.8.0-rc4 Redis/redis-For-Windows:2.8.0-rc3 Redis/redis-For-Windows:2.8.0-rc2 Redis/redis-For-Windows:2.8.0-rc1 Redis/redis-For-Windows:2.6.14 Redis/redis-For-Windows:Win2.6.8 Redis/redis-For-Windows:2.6.13 Redis/redis-For-Windows:2.6.12 Redis/redis-For-Windows:2.6.11 Redis/redis-For-Windows:2.6.10 Redis/redis-For-Windows:2.6.9 Redis/redis-For-Windows:2.6.8 Redis/redis-For-Windows:2.6.7 Redis/redis-For-Windows:2.6.6 Redis/redis-For-Windows:2.6.5 Redis/redis-For-Windows:2.6.4 Redis/redis-For-Windows:2.6.3 Redis/redis-For-Windows:2.6.2 Redis/redis-For-Windows:2.6.1 Redis/redis-For-Windows:2.6.0 Redis/redis-For-Windows:2.6.0-rc8 Redis/redis-For-Windows:2.6.0-rc7 Redis/redis-For-Windows:2.6.0-rc6 Redis/redis-For-Windows:2.6.0-rc5 Redis/redis-For-Windows:2.6.0-rc4 Redis/redis-For-Windows:2.6.0-rc3 Redis/redis-For-Windows:2.6.0-rc2 Redis/redis-For-Windows:2.6.0-rc1 Redis/redis-For-Windows:2.4.11 Redis/redis-For-Windows:2.4.10 Redis/redis-For-Windows:2.4.9 Redis/redis-For-Windows:2.4.8 Redis/redis-For-Windows:2.4.6_win32 Redis/redis-For-Windows:2.4.7 Redis/redis-For-Windows:2.4.6 Redis/redis-For-Windows:2.4.5 Redis/redis-For-Windows:2.4.4 Redis/redis-For-Windows:2.4.3 Redis/redis-For-Windows:2.4.2 Redis/redis-For-Windows:2.4.1 Redis/redis-For-Windows:2.4.0 Redis/redis-For-Windows:2.4.0-rc8 Redis/redis-For-Windows:2.4.0-rc7 Redis/redis-For-Windows:2.4.0-rc6 Redis/redis-For-Windows:2.4.0-rc5 Redis/redis-For-Windows:2.4.0-rc4 Redis/redis-For-Windows:2.4.0-rc3 Redis/redis-For-Windows:with-deprecated-diskstore Redis/redis-For-Windows:2.4.0-rc2 Redis/redis-For-Windows:2.4.0-rc1 Redis/redis-For-Windows:2.2.4 Redis/redis-For-Windows:2.2.3 Redis/redis-For-Windows:3.0-alpha0 Redis/redis-For-Windows:2.2.2 Redis/redis-For-Windows:2.2.1 Redis/redis-For-Windows:2.2.0 Redis/redis-For-Windows:2.2.0-rc4 Redis/redis-For-Windows:2.3-alpha0 Redis/redis-For-Windows:2.2.0-rc3 Redis/redis-For-Windows:2.2.0-rc2 Redis/redis-For-Windows:2.2.0-rc1 Redis/redis-For-Windows:2.2-alpha6 Redis/redis-For-Windows:2.2-alpha5 Redis/redis-For-Windows:2.2-alpha4 Redis/redis-For-Windows:2.2-alpha3 Redis/redis-For-Windows:2.2-alpha2 Redis/redis-For-Windows:2.2-alpha1 Redis/redis-For-Windows:2.2-alpha0 Redis/redis-For-Windows:v2.1.1-watch Redis/redis-For-Windows:v2.0.0-rc1 Redis/redis-For-Windows:v1.3.12 Redis/redis-For-Windows:v1.3.11 Redis/redis-For-Windows:v1.3.10 Redis/redis-For-Windows:v1.3.9 Redis/redis-For-Windows:v1.3.8 Redis/redis-For-Windows:v1.3.7 Redis/redis-For-Windows:1.3.6 Redis/redis-For-Windows:vm-playpen
..
compare: Redis/redis-For-Windows:3.0.4
Branches Tags
Redis/redis-For-Windows:win-5.0 Redis/redis-For-Windows:develop Redis/redis-For-Windows:win-6.2
Redis/redis-For-Windows:5.0.14.1 Redis/redis-For-Windows:v5.0.14.1 Redis/redis-For-Windows:v5.0.14 Redis/redis-For-Windows:v5.0.10 Redis/redis-For-Windows:v5.0.9 Redis/redis-For-Windows:v4.0.14.2 Redis/redis-For-Windows:v4.0.14.1 Redis/redis-For-Windows:v4.0.14 Redis/redis-For-Windows:v4.0.2.3-alpha Redis/redis-For-Windows:v4.0.2.2-alpha Redis/redis-For-Windows:v4.0.2.1-alpha Redis/redis-For-Windows:v4.0.2-alpha Redis/redis-For-Windows:win-3.2.100 Redis/redis-For-Windows:win-3.0.504 Redis/redis-For-Windows:win-3.0.503 Redis/redis-For-Windows:win-2.8.2402 Redis/redis-For-Windows:win-3.0.502 Redis/redis-For-Windows:win-2.8.2401 Redis/redis-For-Windows:3.2.1 Redis/redis-For-Windows:win-3.2.0000-preview Redis/redis-For-Windows:3.2.0 Redis/redis-For-Windows:3.2.0-rc3 Redis/redis-For-Windows:3.2.0-rc2 Redis/redis-For-Windows:win-2.8.2400 Redis/redis-For-Windows:win-3.0.501 Redis/redis-For-Windows:3.2-rc1 Redis/redis-For-Windows:3.0.6 Redis/redis-For-Windows:2.8.24 Redis/redis-For-Windows:win-3.0.500 Redis/redis-For-Windows:win-3.0.500-rc2 Redis/redis-For-Windows:win-3.0.500-rc1 Redis/redis-For-Windows:win-2.8.2104 Redis/redis-For-Windows:3.0.5 Redis/redis-For-Windows:2.8.23 Redis/redis-For-Windows:win-3.0.300-beta1 Redis/redis-For-Windows:win-2.8.2103 Redis/redis-For-Windows:2.8.22 Redis/redis-For-Windows:3.0.4 Redis/redis-For-Windows:win-3.0.300-alpha3 Redis/redis-For-Windows:win-2.8.2102 Redis/redis-For-Windows:win-3.0.300-alpha2 Redis/redis-For-Windows:win-3.0.100-alpha1 Redis/redis-For-Windows:3.0.3 Redis/redis-For-Windows:win-2.8.2101 Redis/redis-For-Windows:win-2.8.21 Redis/redis-For-Windows:3.0.2 Redis/redis-For-Windows:2.8.21 Redis/redis-For-Windows:2.8.20 Redis/redis-For-Windows:3.0.1 Redis/redis-For-Windows:win-2.8.19.1 Redis/redis-For-Windows:3.0.0 Redis/redis-For-Windows:3.0.0-rc6 Redis/redis-For-Windows:3.0.0-rc5 Redis/redis-For-Windows:win-2.8.19-rc1 Redis/redis-For-Windows:win-2.8.19 Redis/redis-For-Windows:3.0.0-rc4 Redis/redis-For-Windows:win-2.8.19-pre2 Redis/redis-For-Windows:win-2.8.17.4 Redis/redis-For-Windows:3.0.0-rc3 Redis/redis-For-Windows:3.0.0-rc2 Redis/redis-For-Windows:win-2.8.17.3 Redis/redis-For-Windows:win-2.8.19-pre1 Redis/redis-For-Windows:win-2.8.17.2 Redis/redis-For-Windows:win-2.8.17.1 Redis/redis-For-Windows:2.8.19 Redis/redis-For-Windows:2.8.18 Redis/redis-For-Windows:3.0.0-rc1 Redis/redis-For-Windows:2.8.17 Redis/redis-For-Windows:2.8.16 Redis/redis-For-Windows:2.8.15 Redis/redis-For-Windows:win2.8.12 Redis/redis-For-Windows:2.8.14 Redis/redis-For-Windows:3.0.0-beta8 Redis/redis-For-Windows:2.8.13 Redis/redis-For-Windows:3.0.0-beta7 Redis/redis-For-Windows:win_2.8.9 Redis/redis-For-Windows:2.8.12 Redis/redis-For-Windows:2.8.11 Redis/redis-For-Windows:3.0.0-beta6 Redis/redis-For-Windows:2.8.10 Redis/redis-For-Windows:3.0.0-beta5 Redis/redis-For-Windows:3.0.0-beta4 Redis/redis-For-Windows:2.6.14_msopentech Redis/redis-For-Windows:2.8.4_msopentech Redis/redis-For-Windows:msopentech_2.8.4 Redis/redis-For-Windows:3.0.0-beta3 Redis/redis-For-Windows:2.8.9 Redis/redis-For-Windows:2.8.8 Redis/redis-For-Windows:3.0.0-beta2 Redis/redis-For-Windows:2.8.7 Redis/redis-For-Windows:2.8.6 Redis/redis-For-Windows:3.0.0-beta1 Redis/redis-For-Windows:2.8.5 Redis/redis-For-Windows:2.8.4 Redis/redis-For-Windows:2.8.3 Redis/redis-For-Windows:2.8.2 Redis/redis-For-Windows:2.8.1 Redis/redis-For-Windows:2.8.0 Redis/redis-For-Windows:2.8.0-rc6 Redis/redis-For-Windows:2.8.0-rc5 Redis/redis-For-Windows:2.8.0-rc4 Redis/redis-For-Windows:2.8.0-rc3 Redis/redis-For-Windows:2.8.0-rc2 Redis/redis-For-Windows:2.8.0-rc1 Redis/redis-For-Windows:2.6.14 Redis/redis-For-Windows:Win2.6.8 Redis/redis-For-Windows:2.6.13 Redis/redis-For-Windows:2.6.12 Redis/redis-For-Windows:2.6.11 Redis/redis-For-Windows:2.6.10 Redis/redis-For-Windows:2.6.9 Redis/redis-For-Windows:2.6.8 Redis/redis-For-Windows:2.6.7 Redis/redis-For-Windows:2.6.6 Redis/redis-For-Windows:2.6.5 Redis/redis-For-Windows:2.6.4 Redis/redis-For-Windows:2.6.3 Redis/redis-For-Windows:2.6.2 Redis/redis-For-Windows:2.6.1 Redis/redis-For-Windows:2.6.0 Redis/redis-For-Windows:2.6.0-rc8 Redis/redis-For-Windows:2.6.0-rc7 Redis/redis-For-Windows:2.6.0-rc6 Redis/redis-For-Windows:2.6.0-rc5 Redis/redis-For-Windows:2.6.0-rc4 Redis/redis-For-Windows:2.6.0-rc3 Redis/redis-For-Windows:2.6.0-rc2 Redis/redis-For-Windows:2.6.0-rc1 Redis/redis-For-Windows:2.4.11 Redis/redis-For-Windows:2.4.10 Redis/redis-For-Windows:2.4.9 Redis/redis-For-Windows:2.4.8 Redis/redis-For-Windows:2.4.6_win32 Redis/redis-For-Windows:2.4.7 Redis/redis-For-Windows:2.4.6 Redis/redis-For-Windows:2.4.5 Redis/redis-For-Windows:2.4.4 Redis/redis-For-Windows:2.4.3 Redis/redis-For-Windows:2.4.2 Redis/redis-For-Windows:2.4.1 Redis/redis-For-Windows:2.4.0 Redis/redis-For-Windows:2.4.0-rc8 Redis/redis-For-Windows:2.4.0-rc7 Redis/redis-For-Windows:2.4.0-rc6 Redis/redis-For-Windows:2.4.0-rc5 Redis/redis-For-Windows:2.4.0-rc4 Redis/redis-For-Windows:2.4.0-rc3 Redis/redis-For-Windows:with-deprecated-diskstore Redis/redis-For-Windows:2.4.0-rc2 Redis/redis-For-Windows:2.4.0-rc1 Redis/redis-For-Windows:2.2.4 Redis/redis-For-Windows:2.2.3 Redis/redis-For-Windows:3.0-alpha0 Redis/redis-For-Windows:2.2.2 Redis/redis-For-Windows:2.2.1 Redis/redis-For-Windows:2.2.0 Redis/redis-For-Windows:2.2.0-rc4 Redis/redis-For-Windows:2.3-alpha0 Redis/redis-For-Windows:2.2.0-rc3 Redis/redis-For-Windows:2.2.0-rc2 Redis/redis-For-Windows:2.2.0-rc1 Redis/redis-For-Windows:2.2-alpha6 Redis/redis-For-Windows:2.2-alpha5 Redis/redis-For-Windows:2.2-alpha4 Redis/redis-For-Windows:2.2-alpha3 Redis/redis-For-Windows:2.2-alpha2 Redis/redis-For-Windows:2.2-alpha1 Redis/redis-For-Windows:2.2-alpha0 Redis/redis-For-Windows:v2.1.1-watch Redis/redis-For-Windows:v2.0.0-rc1 Redis/redis-For-Windows:v1.3.12 Redis/redis-For-Windows:v1.3.11 Redis/redis-For-Windows:v1.3.10 Redis/redis-For-Windows:v1.3.9 Redis/redis-For-Windows:v1.3.8 Redis/redis-For-Windows:v1.3.7 Redis/redis-For-Windows:1.3.6 Redis/redis-For-Windows:vm-playpen
67 Commits
3.0.1 .. 3.0.4
Author SHA1 Message Date
antirez 4698284b41 Redis 3.0.4. 2015-09-08 10:02:10 +02:00
antirez 718a826c51 Sentinel: command arity check added where missing. 2015-09-08 09:33:39 +02:00
Rogerio Goncalves 3915e1c71a Check args before run ckquorum. Fix issue #2635 2015-09-08 09:28:25 +02:00
antirez ce4c17308e Fix merge issues in 490847c. 2015-09-07 17:29:21 +02:00
antirez 490847c681 Undo slaves state change on failed rdbSaveToSlavesSockets(). 
As Oran Agra suggested, in startBgsaveForReplication() when the BGSAVE
attempt returns an error, we scan the list of slaves in order to remove
them since there is no way to serve them currently.

However we check for the replication state BGSAVE_START, which was
modified by rdbSaveToSlaveSockets() before forking(). So when fork fails
the state of slaves remain BGSAVE_END and no cleanup is performed.

This commit fixes the problem by making rdbSaveToSlavesSockets() able to
undo the state change on fork failure.
 2015-09-07 16:21:24 +02:00
antirez c20218eb57 Sentinel: fix bug in config rewriting during failover 
We have a check to rewrite the config properly when a failover is in
progress, in order to add the current (already failed over) master as
slave, and don't include in the slave list the promoted slave itself.

However there was an issue, the variable with the right address was
computed but never used when the code was modified, and no tests are
available for this feature for two reasons:

1. The Sentinel unit test currently does not test Sentinel ability to
persist its state at all.
2. It is a very hard to trigger state since it lasts for little time in
the context of the testing framework.

However this feature should be covered in the test in some way.

The bug was found by @badboy using the clang static analyzer.

Effects of the bug on safety of Sentinel
===

This bug results in severe issues in the following case:

1. A Sentinel is elected leader.
2. During the failover, it persists a wrong config with a known-slave
entry listing the master address.
3. The Sentinel crashes and restarts, reading invalid configuration from
disk.
4. It sees that the slave now does not obey the logical configuration
(should replicate from the current master), so it sends a SLAVEOF
command to the master (since the slave master is the same) creating a
replication loop (attempt to replicate from itself) which Redis is
currently unable to detect.
5. This means that the master is no longer available because of the bug.

However the lack of availability should be only transient (at least
in my tests, but other states could be possible where the problem
is not recovered automatically) because:

6. Sentinels treat masters reporting to be slaves as failing.
7. A new failover is triggered, and a slave is promoted to master.

Bug lifetime
===

The bug is there forever. Commit 16237d78 actually tried to fix the bug
but in the wrong way (the computed variable was never used! My fault).
So this bug is there basically since the start of Sentinel.

Since the bug is hard to trigger, I remember little reports matching
this condition, but I remember at least a few. Also in automated tests
where instances were stopped and restarted multiple times automatically
I remember hitting this issue, however I was not able to reproduce nor
to determine with the information I had at the time what was causing the
issue.
 2015-09-07 15:52:44 +02:00
antirez 34d87be519 Sentinel: clarify effect of resetting failover_start_time. 2015-09-07 15:52:33 +02:00
ubuntu 0513de624c SCAN iter parsing changed from atoi to chartoull 2015-09-07 13:25:30 +02:00
antirez 49f2f691cb Log client details on SLAVEOF command having an effect. 2015-08-21 15:30:49 +02:00
antirez c2ff9de31b startBgsaveForReplication(): handle waiting slaves state change. 
Before this commit, after triggering a BGSAVE it was up to the caller of
startBgsavForReplication() to handle slaves in WAIT_BGSAVE_START in
order to update them accordingly. However when the replication target is
the socket, this is not possible since the process of updating the
slaves and sending the FULLRESYNC reply must be coupled with the process
of starting an RDB save (the reason is, we need to send the FULLSYNC
command and spawn a child that will start to send RDB data to the slaves
ASAP).

This commit moves the responsibility of handling slaves in
WAIT_BGSAVE_START to startBgsavForReplication() so that for both
diskless and disk-based replication we have the same chain of
responsiblity. In order accomodate such change, the syncCommand() also
needs to put the client in the slave list ASAP (just after the initial
checks) and not at the end, so that startBgsavForReplication() can find
the new slave alrady in the list.

Another related change is what happens if the BGSAVE fails because of
fork() or other errors: we now remove the slave from the list of slaves
and send an error, scheduling the slave connection to be terminated.

As a side effect of this change the following errors found by
Oran Agra are fixed (thanks!):

1. rdbSaveToSlavesSockets() on failed fork will get the slaves cleaned
up, otherwise they remain in a wrong state forever since we setup them
for full resync before actually trying to fork.

2. updateSlavesWaitingBgsave() with replication target set as "socket"
was broken since the function changed the slaves state from
WAIT_BGSAVE_START to WAIT_BGSAVE_END via
replicationSetupSlaveForFullResync(), so later rdbSaveToSlavesSockets()
will not find any slave in the right state (WAIT_BGSAVE_START) to feed.
 2015-08-21 11:55:14 +02:00
antirez c5a8c8e907 Fixed issues introduced during last merge. 2015-08-20 10:21:50 +02:00
antirez 4363fa1d76 Force slaves to resync after unsuccessful PSYNC. 
Using chained replication where C is slave of B which is in turn slave of
A, if B reconnects the replication link with A but discovers it is no
longer possible to PSYNC, slaves of B must be disconnected and PSYNC
not allowed, since the new B dataset may be completely different after
the synchronization with the master.

Note that there are varius semantical differences in the way this is
handled now compared to the past. In the past the semantics was:

1. When a slave lost connection with its master, disconnected the chained
slaves ASAP. Which is not needed since after a successful PSYNC with the
master, the slaves can continue and don't need to resync in turn.

2. However after a failed PSYNC the replication backlog was not reset, so a
slave was able to PSYNC successfully even if the instance did a full
sync with its master, containing now an entirely different data set.

Now instead chained slaves are not disconnected when the slave lose the
connection with its master, but only when it is forced to full SYNC with
its master. This means that if the slave having chained slaves does a
successful PSYNC all its slaves can continue without troubles.

See issue #2694 for more details.
 2015-08-20 10:19:42 +02:00
antirez 5a9bc7cc10 replicationHandleMasterDisconnection() belongs to replication.c. 2015-08-20 10:19:10 +02:00
antirez 0b76524983 flushSlavesOutputBuffers(): details clarified via comments. 
Talking with @oranagra we had to reason a little bit to understand if
this function could ever flush the output buffers of the wrong slaves,
having online state but actually not being ready to receive writes
before the first ACK is received from them (this happens with diskless
replication).

Next time we'll just read this comment.
 2015-08-20 10:15:13 +02:00
antirez 421582f845 checkTcpBacklogSetting() now called in Sentinel mode too. 2015-08-20 10:14:48 +02:00
antirez 3cfca5afdc slaveTryPartialResynchronization and syncWithMaster: better synergy. 
It is simpler if removing the read event handler from the FD is up to
slaveTryPartialResynchronization, after all it is only called in the
context of syncWithMaster.

This commit also makes sure that on error all the event handlers are
removed from the socket before closing it.
 2015-08-07 12:20:03 +02:00
antirez 0643ba066e syncWithMaster(): non blocking state machine. 2015-08-07 12:20:03 +02:00
antirez 809e2f4da8 startBgsaveForReplication(): log what you really do. 2015-08-06 12:34:25 +02:00
antirez ba3237604f Replication: add REPLCONF CAPA EOF support. 
Add the concept of slaves capabilities to Redis, the slave now presents
to the Redis master with a set of capabilities in the form:

    REPLCONF capa SOMECAPA capa OTHERCAPA ...

This has the effect of setting slave->slave_capa with the corresponding
SLAVE_CAPA macros that the master can test later to understand if it
the slave will understand certain formats and protocols of the
replication process. This makes it much simpler to introduce new
replication capabilities in the future in a way that don't break old
slaves or masters.

This patch was designed and implemented together with Oran Agra
(@oranagra).
 2015-08-06 12:33:52 +02:00
antirez ade9bf7cb3 Fix synchronous readline "\n" handling. 
Our function to read a line with a timeout handles newlines as requests
to refresh the timeout, however the code kept subtracting the buffer
size left every time a newline was received, for a bug in the loop
logic. Fixed by this commit.
 2015-08-05 16:59:23 +02:00
antirez 64b28d1097 Fix replication slave pings period. 
For PINGs we use the period configured by the user, but for the newlines
of slaves waiting for an RDB to be created (including slaves waiting for
the FULLRESYNC reply) we need to ping with frequency of 1 second, since
the timeout is fixed and needs to be refreshed.
 2015-08-05 16:59:16 +02:00
antirez dcc5ee156f Fix RDB encoding test for new csvdump format. 2015-08-05 14:05:39 +02:00
antirez 58a8c0626c Remove slave state change handled by replicationSetupSlaveForFullResync(). 2015-08-05 13:59:22 +02:00
antirez 27c0ab238b Make sure we re-emit SELECT after each new slave full sync setup. 
In previous commits we moved the FULLRESYNC to the moment we start the
BGSAVE, so that the offset we provide is the right one. However this
also means that we need to re-emit the SELECT statement every time a new
slave starts to accumulate the changes.

To obtian this effect in a more clean way, the function that sends the
FULLRESYNC reply was overloaded with a more important role of also doing
this and chanigng the slave state. So it was renamed to
replicationSetupSlaveForFullResync() to better reflect what it does now.
 2015-08-05 13:57:38 +02:00
antirez 547ccc4b5e Test: csvdump now scans all DBs. 2015-08-05 13:53:32 +02:00
antirez 99e4cf4d84 Don't send SELECT to slaves in WAIT_BGSAVE_START state. 2015-08-05 13:53:28 +02:00
antirez 1dccb6cffb PSYNC test: also test the vanilla SYNC. 2015-08-05 13:52:57 +02:00
antirez f6c65d3f45 syncCommand() comments improved. 2015-08-05 13:52:50 +02:00
antirez 9d2972183e PSYNC initial offset fix. 
This commit attempts to fix a bug involving PSYNC and diskless
replication (currently experimental) found by Yuval Inbar from Redis Labs
and that was later found to have even more far reaching effects (the bug also
exists when diskstore is off).

The gist of the bug is that, a Redis master replies with +FULLRESYNC to
a PSYNC attempt that fails and requires a full resynchronization.
However, the baseline offset sent along with FULLRESYNC was always the
current master replication offset. This is not ok, because there are
many reasosn that may delay the RDB file creation. And... guess what,
the master offset we communicate must be the one of the time the RDB
was created. So for example:

1) When the BGSAVE for replication is delayed since there is one
   already but is not good for replication.
2) When the BGSAVE is not needed as we attach one currently ongoing.
3) When because of diskless replication the BGSAVE is delayed.

In all the above cases the PSYNC reply is wrong and the slave may
reconnect later claiming to need a wrong offset: this may cause
data curruption later.
 2015-08-05 13:51:43 +02:00
antirez 32e979801b Test PSYNC with diskless replication. 
Thanks to Oran Agra from Redis Labs for providing this patch.
 2015-08-05 13:45:02 +02:00
antirez f15bf6c8f8 Redis 3.0.3. 2015-07-17 11:50:21 +02:00
Jan-Erik Rediger 9b0a47cbc8 Do not attempt to lock on Solaris 2015-07-17 11:04:24 +02:00
MOON_CLJ f22a9c0f78 pfcount support multi keys 2015-07-17 10:56:28 +02:00
Yongyue Sun a57aa8831e bugfix: errno might change before logging 
Signed-off-by: Yongyue Sun <abioy.sun@gmail.com>
 2015-07-17 10:47:26 +02:00
Tom Kiemes ad0082fde4 Fix: aof_delayed_fsync is not reset 
aof_delayed_fsync was not set to 0 when calling CONFIG RESETSTAT
 2015-07-17 10:39:30 +02:00
antirez b029ff11b6 Client timeout handling improved. 
The previos attempt to process each client at least once every ten
seconds was not a good idea, because:

1. Usually because of the past min iterations set to 50, you get much
better processing period most of the times.

2. However when there are many clients and a normal setting for
server.hz, the edge case is triggered, and waiting 10 seconds for a
BLPOP that asked for 1 second is not ok.

3. Moreover, because of the high min-itereations limit of 50, when HZ
was set to an high value, the actual behavior was to process a lot of
clients per second.

Also the function checking for timeouts called gettimeofday() at each
iteration which can be costly.

The new implementation will try to process each client once per second,
gets the current time as argument, and does not attempt to process more
than 5 clients per iteration if not needed.

So now:

1. The CPU usage of an idle Redis process is the same or better.
2. The CPU usage of a busy Redis process is the same or better.
3. However a non trivial amount of work may be performed per iteration
when there are many many clients. In this particular case the user may
want to raise the "HZ" value if needed.

Btw with 4000 clients it was still not possible to noticy any actual
latency created by processing 400 clients per second, since the work
performed for each client is pretty small.
 2015-07-16 11:03:40 +02:00
antirez 5dcba26b31 Clarify a comment in clientsCron(). 2015-07-16 11:03:40 +02:00
antirez 7ae1d4d6f5 EXISTS is now variadic. 
The new return value is the number of keys existing, among the ones
specified in the command line, counting the same key multiple times if
given multiple times (and if it exists).

See PR #2667.
 2015-07-13 18:23:18 +02:00
antirez 55e8d4cf1b Add 3.0 changed config default for maxmemory policy to release notes. 2015-06-29 17:23:52 +02:00
linfangrong d4a7c9e1ab Update t_zset.c 2015-06-11 15:15:45 +02:00
antirez 4e8759c65e Use best effort address binding to connect to the master 
We usually want to reach the master using the address of the interface
Redis is bound to (via the "bind" config option). That's useful since
the master will get (and publish) the slave address getting the peer
name of the incoming socket connection from the slave.

However, when this is not possible, for example because the slave is
bound to the loopback interface but repliaces from a master accessed via
an external interface, we want to still connect with the master even
from a different interface: in this case it is not really important that
the master will provide any other address, while it is vital to be able
to replicate correctly.

Related to issues #2609 and #2612.
 2015-06-11 13:01:06 +02:00
antirez a8d7f00e2e anet.c: new API anetTcpNonBlockBestEffortBindConnect() 
This performs a best effort source address binding attempt. If it is
possible to bind the local address and still have a successful
connect(), then this socket is returned. Otherwise the call is retried
without source address binding attempt.

Related to issues #2609 and #2612.
 2015-06-11 13:01:04 +02:00
antirez af8a9de663 anetTcpGenericConnect(), jump to error not end on error 
Two code paths jumped to the "ok, return the socket to the user" code
path to handle error conditions.

Related to issues #2609 and #2612.
 2015-06-11 13:00:55 +02:00
antirez d815289d54 Don't try to bind the source address for MIGRATE 
Related to issues #2609 and #2612.
 2015-06-11 13:00:52 +02:00
antirez f58d67b015 Fix 3.0.2 release notes to give full credits. 2015-06-04 11:59:45 +02:00
antirez 01888d1e58 Redis 3.0.2 2015-06-04 11:35:56 +02:00
Ben Murphy 30278061cc hide access to debug table 2015-06-03 13:36:00 +02:00
Ben Murphy 49efe300af disable loading lua bytecode 2015-06-03 13:35:59 +02:00
antirez 702f914771 Scripting: Lua cmsgpack lib updated to include str8 support 2015-06-03 08:45:58 +02:00
Itamar Haber e0b2e24830 Removed incorrect suggestion 
DEL/INCR/DECR and others could be NTH but apparently never made it to the implementation of SORT
 2015-05-29 12:27:59 +02:00
antirez 0da453160b Test: ZADD CH tests 2015-05-29 12:27:17 +02:00
antirez 1d8973c47d ZADD RETCH option renamed CH 
From Twitter:

    "@antirez that’s an awfully-named command :(
     http://en.wikipedia.org/wiki/Retching"
 2015-05-29 12:27:17 +02:00
antirez df7add9e70 Test: ZADD INCR test 2015-05-29 12:27:17 +02:00
antirez 9003483d43 Test: ZADD NX and XX options tests 2015-05-29 12:27:17 +02:00
antirez a13d6378c1 ZADD RETCH option: Return number of elements added or updated 
Normally ZADD only returns the number of elements added to a sorted
set, using the RETCH option it returns the sum of elements added or
for which the score was updated.
 2015-05-29 12:27:17 +02:00
antirez fa17e2daf0 ZADD NX and XX options 2015-05-29 12:27:17 +02:00
antirez ff7c1faa12 ZADD implemenation able to take options. 2015-05-29 12:27:16 +02:00
therealbill e252e9f231 adding a sentinel command: "flushconfig" 
This new command triggers a config flush to save the in-memory config to
disk. This is useful for cases of a configuration management system or a
package manager wiping out your sentinel config while the process is
still running - and has not yet been restarted. It can also be useful
for scripting a backup and migrate or clone of a running sentinel.
 2015-05-25 12:06:52 +02:00
antirez 9c0a68861e Sentinel: CKQUORUM tests 2015-05-19 12:27:19 +02:00
antirez 5844f5d0d1 Sentinel: SENTINEL CKQUORUM command 
A way for monitoring systems to check that Sentinel is technically able
to reach the quorum and failover, using the currently visible Sentinels.
 2015-05-19 12:27:19 +02:00
antirez a97658f293 Merge branch '3.0' of github.com:/antirez/redis into 3.0 2015-05-15 17:40:03 +02:00
antirez 138e7af57b Rewrite smoveCommand test with ternary operator 2015-05-15 17:39:37 +02:00
Glenn Nethercutt 3a9f41ad86 uphold the smove contract to return 0 when the element is not a member of the source set, even if source=dest 2015-05-15 17:39:37 +02:00
Salvatore Sanfilippo 71fba427c2 Merge pull request #2567 from ryanschwartz/patch-1 
Update 00-RELEASENOTES
 2015-05-15 17:22:21 +02:00
Jungtaek Lim 08d4df8d31 protocol error log should be seen debug/verbose level 2015-05-15 17:06:42 +02:00
Ryan Schwartz e7422ef166 Update 00-RELEASENOTES 
Fix typo in 00-RELEASENOTES.
 2015-05-07 14:28:21 -05:00
antirez 362032e43a Release notes: Sentinel upgrade urgency moved to moderate 2015-05-05 11:16:20 +02:00
26 changed files with 1029 additions and 361 deletions
Expand all files Collapse all files
00-RELEASENOTES
+91 -2
View File
@@ -10,9 +10,93 @@ HIGH: There is a critical bug that may affect a subset of users. Upgrade!
CRITICAL: There is a critical bug affecting MOST USERS. Upgrade ASAP.
--------------------------------------------------------------------------------
--[ Redis 3.0.4 ] Release date: 8 Sep 2015
Upgrade urgency: HIGH for Redis and Sentinel. However note that in order to
fix certain replication bugs, the replication internals were
modified in a very heavy way. So while this release is
conceptually saner, it may contain regressions. For this
reason, before the release, QA activities were performed by
me (antirez) and Redis Labs and no evident bug was found.
* [FIX] A number of bugs related to replication PSYNC and the (yet experimental)
diskless replication feature were fixed. The bugs could lead to
inconsistency between masters and slaves. (Salvatore Sanfilippo, Oran
Agra fixed the issue found by Yuval Inbar)
* [FIX] A replication bug in the context of PSYNC partial resynchonization was
found and fixed. This bug happens even when diskless replication is off
in the case different slaves connect at different times while the master
is creating an RDB file, and later a partial resynchronization is
attempted by a slave that connected not as the first one. (Salvatore
Sanfilippo, Oran Agra)
* [FIX] Chained replication and PSYNC interactions leading to potential stale
chained slaves data set, see issue #2694. (Salvatore Sanfilippo fixed
an issue reported by "GeorgeBJ" user at Github)
* [FIX] redis-cli --scan iteration fixed when returned cursor overflows
32 bit signed integer. (Ofir Luzon, Yuval Inbar)
* [FIX] Senitnel: fixed a bug during the master switch process, where for a
failed conditional check, the new configuration is rewritten, during
a small window of time, in a corrupted way where the master is
also reported to be one of the slaves. This bug is rare to trigger
but apparently it happens in the wild, and the effect is to see
a replication loop where the master will try to replicate with itself.
A detailed explanation of the bug and its effects can be found in
the commit message here: https://github.com/antirez/redis/commit/c20218eb5770b2cafb12bc7092313b8358fedc0a.
The bug was found by Jan-Erik Rediger using a static analyzer and
fixed by Salvatore Sanfilippo.
* [FIX] Sentinel lack of arity checks for certain commands.
(Rogerio Goncalves, Salvatore Sanfilippo)
* [NEW] Replication internals rewritten in order to be more resistant to bugs.
The replication handshake in the slave side was rewritten as a non
blocking state machine. (Salvatore Sanfilippo, Oran Agra)
* [NEW] New "replication capabilities" feature introduced in order to signal
from the master to the slave what are the features supported, so that
the master can choose the kind of replication to start (diskless or
not) when master and slave are of different versions. (Oran Agra,
Salvatore Sanfilippo)
* [NEW] Log clients details when SLAVEOF command is received. (Salvatore
Sanfilippo with inputs from Nick Craver and Marc Gravell).
--[ Redis 3.0.3 ] Release date: 17 Jul 2015
Upgrade urgency: LOW for Redis and Sentinel.
* [FIX] Fix blocking operations timeout precision when HZ is at its default
value (not increased) and there are thousands of clients connected
at the same time. This bug affected Sidekiq users that experienced
a very long delay for BLPOP and similar commands to return for
timeout. Check commit b029ff1 for more info. (Salvatore Sanfilippo)
* [FIX] MIGRATE "creating socket: Invalid argument" error fix. Check
issues #2609 and #2612 for more info. (Salvatore Sanfilippo)
* [FIX] Be able to connect to the master even when the slave is bound to
just the loopback interface and has no valid public address in the
network the master is reacahble. (Salvatore Sanfilippo)
* [FIX] ZADD with options encoding promotion fixed. (linfangrong)
* [FIX] Reset aof_delayed_fsync on CONFIG RESETSTATS. (Tom Kiemes)
* [FIX] PFCOUNT key parsing in cluster fixed. (MOON_CLJ)
* [FIX] Fix Solaris compilation of Redis 3.0. (Jan-Erik Rediger)
* [NEW] Variadic EXISTS command. Now the command accepts multiple arguments
and returns the total count of existing keys.
--[ Redis 3.0.2 ] Release date: 4 Jun 2015
Upgrade urgency: HIGH for Redis because of a security issue.
LOW for Sentinel.
* [FIX] Critical security issue fix by Ben Murphy: http://t.co/LpGTyZmfS7
* [FIX] SMOVE reply fixed when src and dst keys are the same. (Glenn Nethercutt)
* [FIX] Lua cmsgpack lib updated to support str8 type. (Sebastian Waisbrot)
* [NEW] ZADD support for options: NX, XX, CH. See new doc at redis.io.
(Salvatore Sanfilippo)
* [NEW] Senitnel: CKQUORUM and FLUSHCONFIG commands back ported.
(Salvatore Sanfilippo and Bill Anderson)
--[ Redis 3.0.1 ] Release date: 5 May 2015
Upgrade urgency: LOW for Redis, Sentinel, Cluster.
Upgrade urgency: LOW for Redis and Cluster, MODERATE for Sentinel.
* [FIX] Sentinel memory leak due to hiredis fixed. (Salvatore Sanfilippo)
* [FIX] Sentinel memory leak on duplicated instance. (Charsyam)
@@ -32,7 +116,7 @@ Upgrade urgency: LOW for Redis, Sentinel, Cluster.
* WAIT command to block waiting for a write to be transmitted to
the specified number of slaves.
* MIGRATE connection caching. Much faster keys migraitons.
* MIGARTE new options COPY and REPLACE.
* MIGRATE new options COPY and REPLACE.
* CLIENT PAUSE command: stop processing client requests for a
specified amount of time.
* BITCOUNT performance improvements.
@@ -594,6 +678,11 @@ non-backward compatible changes introduced in the 3.0 release:
'?' is actually the role of the instance. M for master, S for slave, C
if this process is a saving child (for RDB/AOF), and X for Sentinel.
* The default maxmemory policy in Redis 3.0 is no longer "volatile-lru" as
it used to be in 2.8, but "noeviction". The policies behavior is the same
(but LRU eviction is much more precise in 3.0), so only the default value
changed. Just make sure to specify in your redis.conf what you mean.
--------------------------------------------------------------------------------
Credits: Where not specified the implementation and design is done by
deps/lua/src/ldo.c
Vendored
+1 -1
View File
@@ -495,7 +495,7 @@ static void f_parser (lua_State *L, void *ud) {
struct SParser *p = cast(struct SParser *, ud);
int c = luaZ_lookahead(p->z);
luaC_checkGC(L);
tf = ((c == LUA_SIGNATURE[0]) ? luaU_undump : luaY_parser)(L, p->z,
tf = (luaY_parser)(L, p->z,
&p->buff, p->name);
cl = luaF_newLclosure(L, tf->nups, hvalue(gt(L)));
cl->l.p = tf;
deps/lua/src/lua_cmsgpack.c
Vendored
+42 -29
View File
@@ -66,7 +66,7 @@
/* Reverse memory bytes if arch is little endian. Given the conceptual
* simplicity of the Lua build system we prefer check for endianess at runtime.
* The performance difference should be acceptable. */
static void memrevifle(void *ptr, size_t len) {
void memrevifle(void *ptr, size_t len) {
unsigned char *p = (unsigned char *)ptr,
*e = (unsigned char *)p+len-1,
aux;
@@ -96,7 +96,7 @@ typedef struct mp_buf {
size_t len, free;
} mp_buf;
static void *mp_realloc(lua_State *L, void *target, size_t osize,size_t nsize) {
void *mp_realloc(lua_State *L, void *target, size_t osize,size_t nsize) {
void *(*local_realloc) (void *, void *, size_t osize, size_t nsize) = NULL;
void *ud;
@@ -105,7 +105,7 @@ static void *mp_realloc(lua_State *L, void *target, size_t osize,size_t nsize) {
return local_realloc(ud, target, osize, nsize);
}
static mp_buf *mp_buf_new(lua_State *L) {
mp_buf *mp_buf_new(lua_State *L) {
mp_buf *buf = NULL;
/* Old size = 0; new size = sizeof(*buf) */
@@ -117,7 +117,7 @@ static mp_buf *mp_buf_new(lua_State *L) {
return buf;
}
static void mp_buf_append(mp_buf *buf, const unsigned char *s, size_t len) {
void mp_buf_append(mp_buf *buf, const unsigned char *s, size_t len) {
if (buf->free < len) {
size_t newlen = buf->len+len;
@@ -153,7 +153,7 @@ typedef struct mp_cur {
int err;
} mp_cur;
static void mp_cur_init(mp_cur *cursor, const unsigned char *s, size_t len) {
void mp_cur_init(mp_cur *cursor, const unsigned char *s, size_t len) {
cursor->p = s;
cursor->left = len;
cursor->err = MP_CUR_ERROR_NONE;
@@ -173,13 +173,17 @@ static void mp_cur_init(mp_cur *cursor, const unsigned char *s, size_t len) {
/* ------------------------- Low level MP encoding -------------------------- */
static void mp_encode_bytes(mp_buf *buf, const unsigned char *s, size_t len) {
void mp_encode_bytes(mp_buf *buf, const unsigned char *s, size_t len) {
unsigned char hdr[5];
int hdrlen;
if (len < 32) {
hdr[0] = 0xa0 | (len&0xff); /* fix raw */
hdrlen = 1;
} else if (len <= 0xff) {
hdr[0] = 0xd9;
hdr[1] = len;
hdrlen = 2;
} else if (len <= 0xffff) {
hdr[0] = 0xda;
hdr[1] = (len&0xff00)>>8;
@@ -198,7 +202,7 @@ static void mp_encode_bytes(mp_buf *buf, const unsigned char *s, size_t len) {
}
/* we assume IEEE 754 internal format for single and double precision floats. */
static void mp_encode_double(mp_buf *buf, double d) {
void mp_encode_double(mp_buf *buf, double d) {
unsigned char b[9];
float f = d;
@@ -216,7 +220,7 @@ static void mp_encode_double(mp_buf *buf, double d) {
}
}
static void mp_encode_int(mp_buf *buf, int64_t n) {
void mp_encode_int(mp_buf *buf, int64_t n) {
unsigned char b[9];
int enclen;
@@ -288,7 +292,7 @@ static void mp_encode_int(mp_buf *buf, int64_t n) {
mp_buf_append(buf,b,enclen);
}
static void mp_encode_array(mp_buf *buf, int64_t n) {
void mp_encode_array(mp_buf *buf, int64_t n) {
unsigned char b[5];
int enclen;
@@ -311,7 +315,7 @@ static void mp_encode_array(mp_buf *buf, int64_t n) {
mp_buf_append(buf,b,enclen);
}
static void mp_encode_map(mp_buf *buf, int64_t n) {
void mp_encode_map(mp_buf *buf, int64_t n) {
unsigned char b[5];
int enclen;
@@ -336,7 +340,7 @@ static void mp_encode_map(mp_buf *buf, int64_t n) {
/* --------------------------- Lua types encoding --------------------------- */
static void mp_encode_lua_string(lua_State *L, mp_buf *buf) {
void mp_encode_lua_string(lua_State *L, mp_buf *buf) {
size_t len;
const char *s;
@@ -344,13 +348,13 @@ static void mp_encode_lua_string(lua_State *L, mp_buf *buf) {
mp_encode_bytes(buf,(const unsigned char*)s,len);
}
static void mp_encode_lua_bool(lua_State *L, mp_buf *buf) {
void mp_encode_lua_bool(lua_State *L, mp_buf *buf) {
unsigned char b = lua_toboolean(L,-1) ? 0xc3 : 0xc2;
mp_buf_append(buf,&b,1);
}
/* Lua 5.3 has a built in 64-bit integer type */
static void mp_encode_lua_integer(lua_State *L, mp_buf *buf) {
void mp_encode_lua_integer(lua_State *L, mp_buf *buf) {
#if (LUA_VERSION_NUM < 503) && BITS_32
lua_Number i = lua_tonumber(L,-1);
#else
@@ -362,7 +366,7 @@ static void mp_encode_lua_integer(lua_State *L, mp_buf *buf) {
/* Lua 5.2 and lower only has 64-bit doubles, so we need to
* detect if the double may be representable as an int
* for Lua < 5.3 */
static void mp_encode_lua_number(lua_State *L, mp_buf *buf) {
void mp_encode_lua_number(lua_State *L, mp_buf *buf) {
lua_Number n = lua_tonumber(L,-1);
if (IS_INT64_EQUIVALENT(n)) {
@@ -372,10 +376,10 @@ static void mp_encode_lua_number(lua_State *L, mp_buf *buf) {
}
}
static void mp_encode_lua_type(lua_State *L, mp_buf *buf, int level);
void mp_encode_lua_type(lua_State *L, mp_buf *buf, int level);
/* Convert a lua table into a message pack list. */
static void mp_encode_lua_table_as_array(lua_State *L, mp_buf *buf, int level) {
void mp_encode_lua_table_as_array(lua_State *L, mp_buf *buf, int level) {
#if LUA_VERSION_NUM < 502
size_t len = lua_objlen(L,-1), j;
#else
@@ -391,7 +395,7 @@ static void mp_encode_lua_table_as_array(lua_State *L, mp_buf *buf, int level) {
}
/* Convert a lua table into a message pack key-value map. */
static void mp_encode_lua_table_as_map(lua_State *L, mp_buf *buf, int level) {
void mp_encode_lua_table_as_map(lua_State *L, mp_buf *buf, int level) {
size_t len = 0;
/* First step: count keys into table. No other way to do it with the
@@ -418,7 +422,7 @@ static void mp_encode_lua_table_as_map(lua_State *L, mp_buf *buf, int level) {
/* Returns true if the Lua table on top of the stack is exclusively composed
* of keys from numerical keys from 1 up to N, with N being the total number
* of elements, without any hole in the middle. */
static int table_is_an_array(lua_State *L) {
int table_is_an_array(lua_State *L) {
int count = 0, max = 0;
#if LUA_VERSION_NUM < 503
lua_Number n;
@@ -461,14 +465,14 @@ static int table_is_an_array(lua_State *L) {
/* If the length operator returns non-zero, that is, there is at least
* an object at key '1', we serialize to message pack list. Otherwise
* we use a map. */
static void mp_encode_lua_table(lua_State *L, mp_buf *buf, int level) {
void mp_encode_lua_table(lua_State *L, mp_buf *buf, int level) {
if (table_is_an_array(L))
mp_encode_lua_table_as_array(L,buf,level);
else
mp_encode_lua_table_as_map(L,buf,level);
}
static void mp_encode_lua_null(lua_State *L, mp_buf *buf) {
void mp_encode_lua_null(lua_State *L, mp_buf *buf) {
unsigned char b[1];
(void)L;
@@ -476,7 +480,7 @@ static void mp_encode_lua_null(lua_State *L, mp_buf *buf) {
mp_buf_append(buf,b,1);
}
static void mp_encode_lua_type(lua_State *L, mp_buf *buf, int level) {
void mp_encode_lua_type(lua_State *L, mp_buf *buf, int level) {
int t = lua_type(L,-1);
/* Limit the encoding of nested tables to a specified maximum depth, so that
@@ -506,7 +510,7 @@ static void mp_encode_lua_type(lua_State *L, mp_buf *buf, int level) {
* Packs all arguments as a stream for multiple upacking later.
* Returns error if no arguments provided.
*/
static int mp_pack(lua_State *L) {
int mp_pack(lua_State *L) {
int nargs = lua_gettop(L);
int i;
mp_buf *buf;
@@ -687,6 +691,15 @@ void mp_decode_to_lua_type(lua_State *L, mp_cur *c) {
mp_cur_consume(c,9);
}
break;
case 0xd9: /* raw 8 */
mp_cur_need(c,2);
{
size_t l = c->p[1];
mp_cur_need(c,2+l);
lua_pushlstring(L,(char*)c->p+2,l);
mp_cur_consume(c,2+l);
}
break;
case 0xda: /* raw 16 */
mp_cur_need(c,3);
{
@@ -773,7 +786,7 @@ void mp_decode_to_lua_type(lua_State *L, mp_cur *c) {
}
}
static int mp_unpack_full(lua_State *L, int limit, int offset) {
int mp_unpack_full(lua_State *L, int limit, int offset) {
size_t len;
const char *s;
mp_cur c;
@@ -826,18 +839,18 @@ static int mp_unpack_full(lua_State *L, int limit, int offset) {
return cnt;
}
static int mp_unpack(lua_State *L) {
int mp_unpack(lua_State *L) {
return mp_unpack_full(L, 0, 0);
}
static int mp_unpack_one(lua_State *L) {
int mp_unpack_one(lua_State *L) {
int offset = luaL_optinteger(L, 2, 0);
/* Variable pop because offset may not exist */
lua_pop(L, lua_gettop(L)-1);
return mp_unpack_full(L, 1, offset);
}
static int mp_unpack_limit(lua_State *L) {
int mp_unpack_limit(lua_State *L) {
int limit = luaL_checkinteger(L, 2);
int offset = luaL_optinteger(L, 3, 0);
/* Variable pop because offset may not exist */
@@ -846,7 +859,7 @@ static int mp_unpack_limit(lua_State *L) {
return mp_unpack_full(L, limit, offset);
}
static int mp_safe(lua_State *L) {
int mp_safe(lua_State *L) {
int argc, err, total_results;
argc = lua_gettop(L);
@@ -869,7 +882,7 @@ static int mp_safe(lua_State *L) {
}
/* -------------------------------------------------------------------------- */
static const struct luaL_Reg cmds[] = {
const struct luaL_Reg cmds[] = {
{"pack", mp_pack},
{"unpack", mp_unpack},
{"unpack_one", mp_unpack_one},
@@ -877,7 +890,7 @@ static const struct luaL_Reg cmds[] = {
{0}
};
static int luaopen_create(lua_State *L) {
int luaopen_create(lua_State *L) {
int i;
/* Manually construct our module table instead of
* relying on _register or _newlib */
src/anet.c
+23 -5
View File
@@ -264,6 +264,7 @@ static int anetCreateSocket(char *err, int domain) {
#define ANET_CONNECT_NONE 0
#define ANET_CONNECT_NONBLOCK 1
#define ANET_CONNECT_BE_BINDING 2 /* Best effort binding. */
static int anetTcpGenericConnect(char *err, char *addr, int port,
char *source_addr, int flags)
{
@@ -295,7 +296,7 @@ static int anetTcpGenericConnect(char *err, char *addr, int port,
if ((rv = getaddrinfo(source_addr, NULL, &hints, &bservinfo)) != 0)
{
anetSetError(err, "%s", gai_strerror(rv));
goto end;
goto error;
}
for (b = bservinfo; b != NULL; b = b->ai_next) {
if (bind(s,b->ai_addr,b->ai_addrlen) != -1) {
@@ -306,7 +307,7 @@ static int anetTcpGenericConnect(char *err, char *addr, int port,
freeaddrinfo(bservinfo);
if (!bound) {
anetSetError(err, "bind: %s", strerror(errno));
goto end;
goto error;
}
}
if (connect(s,p->ai_addr,p->ai_addrlen) == -1) {
@@ -331,9 +332,17 @@ error:
close(s);
s = ANET_ERR;
}
end:
freeaddrinfo(servinfo);
return s;
/* Handle best effort binding: if a binding address was used, but it is
* not possible to create a socket, try again without a binding address. */
if (s == ANET_ERR && source_addr && (flags & ANET_CONNECT_BE_BINDING)) {
return anetTcpGenericConnect(err,addr,port,NULL,flags);
} else {
return s;
}
}
int anetTcpConnect(char *err, char *addr, int port)
@@ -346,9 +355,18 @@ int anetTcpNonBlockConnect(char *err, char *addr, int port)
return anetTcpGenericConnect(err,addr,port,NULL,ANET_CONNECT_NONBLOCK);
}
int anetTcpNonBlockBindConnect(char *err, char *addr, int port, char *source_addr)
int anetTcpNonBlockBindConnect(char *err, char *addr, int port,
char *source_addr)
{
return anetTcpGenericConnect(err,addr,port,source_addr,ANET_CONNECT_NONBLOCK);
return anetTcpGenericConnect(err,addr,port,source_addr,
ANET_CONNECT_NONBLOCK);
}
int anetTcpNonBlockBestEffortBindConnect(char *err, char *addr, int port,
char *source_addr)
{
return anetTcpGenericConnect(err,addr,port,source_addr,
ANET_CONNECT_NONBLOCK|ANET_CONNECT_BE_BINDING);
}
int anetUnixGenericConnect(char *err, char *path, int flags)
src/anet.h
+1
View File
@@ -50,6 +50,7 @@
int anetTcpConnect(char *err, char *addr, int port);
int anetTcpNonBlockConnect(char *err, char *addr, int port);
int anetTcpNonBlockBindConnect(char *err, char *addr, int port, char *source_addr);
int anetTcpNonBlockBestEffortBindConnect(char *err, char *addr, int port, char *source_addr);
int anetUnixConnect(char *err, char *path);
int anetUnixNonBlockConnect(char *err, char *path);
int anetRead(int fd, char *buf, int count);
src/aof.c
+1 -1
View File
@@ -1169,9 +1169,9 @@ int rewriteAppendOnlyFile(char *filename) {
return REDIS_OK;
werr:
redisLog(REDIS_WARNING,"Write error writing append only file on disk: %s", strerror(errno));
fclose(fp);
unlink(tmpfile);
redisLog(REDIS_WARNING,"Write error writing append only file on disk: %s", strerror(errno));
if (di) dictReleaseIterator(di);
return REDIS_ERR;
}
src/cluster.c
+9 -2
View File
@@ -358,6 +358,11 @@ void clusterSaveConfigOrDie(int do_fsync) {
* On success REDIS_OK is returned, otherwise an error is logged and
* the function returns REDIS_ERR to signal a lock was not acquired. */
int clusterLockConfig(char *filename) {
/* flock() does not exist on Solaris
* and a fcntl-based solution won't help, as we constantly re-open that file,
* which will release _all_ locks anyway
*/
#if !defined(__sun)
/* To lock it, we need to open the file in a way it is created if
* it does not exist, otherwise there is a race condition with other
* processes. */
@@ -385,6 +390,8 @@ int clusterLockConfig(char *filename) {
}
/* Lock acquired: leak the 'fd' by not closing it, so that we'll retain the
* lock to the file as long as the process exists. */
#endif /* __sun */
return REDIS_OK;
}
@@ -4486,8 +4493,8 @@ migrateCachedSocket* migrateGetSocket(redisClient *c, robj *host, robj *port, lo
}
/* Create the socket */
fd = anetTcpNonBlockBindConnect(server.neterr,c->argv[1]->ptr,
atoi(c->argv[2]->ptr),REDIS_BIND_ADDR);
fd = anetTcpNonBlockConnect(server.neterr,c->argv[1]->ptr,
atoi(c->argv[2]->ptr));
if (fd == -1) {
sdsfree(name);
addReplyErrorFormat(c,"Can't connect to target node: %s",
src/db.c
+9 -5
View File
@@ -293,13 +293,17 @@ void delCommand(redisClient *c) {
addReplyLongLong(c,deleted);
}
/* EXISTS key1 key2 ... key_N.
* Return value is the number of keys existing. */
void existsCommand(redisClient *c) {
expireIfNeeded(c->db,c->argv[1]);
if (dbExists(c->db,c->argv[1])) {
addReply(c, shared.cone);
} else {
addReply(c, shared.czero);
long long count = 0;
int j;
for (j = 1; j < c->argc; j++) {
expireIfNeeded(c->db,c->argv[j]);
if (dbExists(c->db,c->argv[j])) count++;
}
addReplyLongLong(c,count);
}
void selectCommand(redisClient *c) {
src/networking.c
+8 -15
View File
@@ -105,6 +105,7 @@ redisClient *createClient(int fd) {
c->repl_ack_off = 0;
c->repl_ack_time = 0;
c->slave_listening_port = 0;
c->slave_capa = SLAVE_CAPA_NONE;
c->reply = listCreate();
c->reply_bytes = 0;
c->obuf_soft_limit_reached_time = 0;
@@ -666,20 +667,6 @@ void disconnectSlaves(void) {
}
}
/* This function is called when the slave lose the connection with the
* master into an unexpected way. */
void replicationHandleMasterDisconnection(void) {
server.master = NULL;
server.repl_state = REDIS_REPL_CONNECT;
server.repl_down_since = server.unixtime;
/* We lost connection with our master, force our slaves to resync
* with us as well to load the new data set.
*
* If server.masterhost is NULL the user called SLAVEOF NO ONE so
* slave resync is not needed. */
if (server.masterhost != NULL) disconnectSlaves();
}
void freeClient(redisClient *c) {
listNode *ln;
@@ -974,7 +961,7 @@ int processInlineBuffer(redisClient *c) {
/* Helper function. Trims query buffer to make the function that processes
* multi bulk requests idempotent. */
static void setProtocolError(redisClient *c, int pos) {
if (server.verbosity >= REDIS_VERBOSE) {
if (server.verbosity <= REDIS_VERBOSE) {
sds client = catClientInfoString(sdsempty(),c);
redisLog(REDIS_VERBOSE,
"Protocol error from client: %s", client);
@@ -1679,6 +1666,12 @@ void flushSlavesOutputBuffers(void) {
redisClient *slave = listNodeValue(ln);
int events;
/* Note that the following will not flush output buffers of slaves
* in STATE_ONLINE but having put_online_on_ack set to true: in this
* case the writable event is never installed, since the purpose
* of put_online_on_ack is to postpone the moment it is installed.
* This is what we want since slaves in this state should not receive
* writes before the first ACK. */
events = aeGetFileEvents(server.el,slave->fd);
if (events & AE_WRITABLE &&
slave->replstate == REDIS_REPL_ONLINE &&
src/rdb.c
+28 -12
View File
@@ -755,9 +755,9 @@ int rdbSave(char *filename) {
return REDIS_OK;
werr:
redisLog(REDIS_WARNING,"Write error saving DB on disk: %s", strerror(errno));
fclose(fp);
unlink(tmpfile);
redisLog(REDIS_WARNING,"Write error saving DB on disk: %s", strerror(errno));
return REDIS_ERR;
}
@@ -1420,7 +1420,7 @@ int rdbSaveToSlavesSockets(void) {
if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_START) {
clientids[numfds] = slave->id;
fds[numfds++] = slave->fd;
slave->replstate = REDIS_REPL_WAIT_BGSAVE_END;
replicationSetupSlaveForFullResync(slave,getPsyncInitialOffset());
/* Put the socket in non-blocking mode to simplify RDB transfer.
* We'll restore it when the children returns (since duped socket
* will share the O_NONBLOCK attribute with the parent). */
@@ -1498,27 +1498,43 @@ int rdbSaveToSlavesSockets(void) {
exitFromChild((retval == REDIS_OK) ? 0 : 1);
} else {
/* Parent */
zfree(clientids); /* Not used by parent. Free ASAP. */
server.stat_fork_time = ustime()-start;
server.stat_fork_rate = (double) zmalloc_used_memory() * 1000000 / server.stat_fork_time / (1024*1024*1024); /* GB per second. */
latencyAddSampleIfNeeded("fork",server.stat_fork_time/1000);
if (childpid == -1) {
redisLog(REDIS_WARNING,"Can't save in background: fork: %s",
strerror(errno));
zfree(fds);
/* Undo the state change. The caller will perform cleanup on
* all the slaves in BGSAVE_START state, but an early call to
* replicationSetupSlaveForFullResync() turned it into BGSAVE_END */
listRewind(server.slaves,&li);
while((ln = listNext(&li))) {
redisClient *slave = ln->value;
int j;
for (j = 0; j < numfds; j++) {
if (slave->id == clientids[j]) {
slave->replstate = REDIS_REPL_WAIT_BGSAVE_START;
break;
}
}
}
close(pipefds[0]);
close(pipefds[1]);
return REDIS_ERR;
} else {
redisLog(REDIS_NOTICE,"Background RDB transfer started by pid %d",
childpid);
server.rdb_save_time_start = time(NULL);
server.rdb_child_pid = childpid;
server.rdb_child_type = REDIS_RDB_CHILD_TYPE_SOCKET;
updateDictResizePolicy();
}
redisLog(REDIS_NOTICE,"Background RDB transfer started by pid %d",childpid);
server.rdb_save_time_start = time(NULL);
server.rdb_child_pid = childpid;
server.rdb_child_type = REDIS_RDB_CHILD_TYPE_SOCKET;
updateDictResizePolicy();
zfree(clientids);
zfree(fds);
return REDIS_OK;
return (childpid == -1) ? REDIS_ERR : REDIS_OK;
}
return REDIS_OK; /* unreached */
return REDIS_OK; /* Unreached. */
}
void saveCommand(redisClient *c) {
src/redis-cli.c
+1 -1
View File
@@ -1582,7 +1582,7 @@ static redisReply *sendScan(unsigned long long *it) {
assert(reply->element[1]->type == REDIS_REPLY_ARRAY);
/* Update iterator */
*it = atoi(reply->element[0]->str);
*it = strtoull(reply->element[0]->str, NULL, 10);
return reply;
}
src/redis.c
+25 -16
View File
@@ -129,7 +129,7 @@ struct redisCommand redisCommandTable[] = {
{"append",appendCommand,3,"wm",0,NULL,1,1,1,0,0},
{"strlen",strlenCommand,2,"rF",0,NULL,1,1,1,0,0},
{"del",delCommand,-2,"w",0,NULL,1,-1,1,0,0},
{"exists",existsCommand,2,"rF",0,NULL,1,1,1,0,0},
{"exists",existsCommand,-2,"rF",0,NULL,1,-1,1,0,0},
{"setbit",setbitCommand,4,"wm",0,NULL,1,1,1,0,0},
{"getbit",getbitCommand,3,"rF",0,NULL,1,1,1,0,0},
{"setrange",setrangeCommand,4,"wm",0,NULL,1,1,1,0,0},
@@ -280,7 +280,7 @@ struct redisCommand redisCommandTable[] = {
{"command",commandCommand,0,"rlt",0,NULL,0,0,0,0,0},
{"pfselftest",pfselftestCommand,1,"r",0,NULL,0,0,0,0,0},
{"pfadd",pfaddCommand,-2,"wmF",0,NULL,1,1,1,0,0},
{"pfcount",pfcountCommand,-2,"r",0,NULL,1,1,1,0,0},
{"pfcount",pfcountCommand,-2,"r",0,NULL,1,-1,1,0,0},
{"pfmerge",pfmergeCommand,-2,"wm",0,NULL,1,-1,1,0,0},
{"pfdebug",pfdebugCommand,-3,"w",0,NULL,0,0,0,0,0},
{"latency",latencyCommand,-2,"arslt",0,NULL,0,0,0,0,0}
@@ -902,9 +902,12 @@ long long getInstantaneousMetric(int metric) {
return sum / REDIS_METRIC_SAMPLES;
}
/* Check for timeouts. Returns non-zero if the client was terminated */
int clientsCronHandleTimeout(redisClient *c) {
time_t now = server.unixtime;
/* Check for timeouts. Returns non-zero if the client was terminated.
* The function gets the current time in milliseconds as argument since
* it gets called multiple times in a loop, so calling gettimeofday() for
* each iteration would be costly without any actual gain. */
int clientsCronHandleTimeout(redisClient *c, mstime_t now_ms) {
time_t now = now_ms/1000;
if (server.maxidletime &&
!(c->flags & REDIS_SLAVE) && /* no timeout for slaves */
@@ -920,7 +923,6 @@ int clientsCronHandleTimeout(redisClient *c) {
/* Blocked OPS timeout is handled with milliseconds resolution.
* However note that the actual resolution is limited by
* server.hz. */
mstime_t now_ms = mstime();
if (c->bpop.timeout != 0 && c->bpop.timeout < now_ms) {
/* Handle blocking operation specific timeout. */
@@ -962,17 +964,23 @@ int clientsCronResizeQueryBuffer(redisClient *c) {
return 0;
}
#define CLIENTS_CRON_MIN_ITERATIONS 5
void clientsCron(void) {
/* Make sure to process at least 1/(server.hz*10) of clients per call.
* Since this function is called server.hz times per second we are sure that
* in the worst case we process all the clients in 10 seconds.
* In normal conditions (a reasonable number of clients) we process
* all the clients in a shorter time. */
/* Make sure to process at least numclients/server.hz of clients
* per call. Since this function is called server.hz times per second
* we are sure that in the worst case we process all the clients in 1
* second. */
int numclients = listLength(server.clients);
int iterations = numclients/(server.hz*10);
int iterations = numclients/server.hz;
mstime_t now = mstime();
/* Process at least a few clients while we are at it, even if we need
* to process less than CLIENTS_CRON_MIN_ITERATIONS to meet our contract
* of processing each client once per second. */
if (iterations < CLIENTS_CRON_MIN_ITERATIONS)
iterations = (numclients < CLIENTS_CRON_MIN_ITERATIONS) ?
numclients : CLIENTS_CRON_MIN_ITERATIONS;
if (iterations < 50)
iterations = (numclients < 50) ? numclients : 50;
while(listLength(server.clients) && iterations--) {
redisClient *c;
listNode *head;
@@ -986,7 +994,7 @@ void clientsCron(void) {
/* The following functions do different service checks on the client.
* The protocol is that they return non-zero if the client was
* terminated. */
if (clientsCronHandleTimeout(c)) continue;
if (clientsCronHandleTimeout(c,now)) continue;
if (clientsCronResizeQueryBuffer(c)) continue;
}
}
@@ -1740,6 +1748,7 @@ void resetServerStats(void) {
}
server.stat_net_input_bytes = 0;
server.stat_net_output_bytes = 0;
server.aof_delayed_fsync = 0;
}
void initServer(void) {
@@ -3651,6 +3660,7 @@ int main(int argc, char **argv) {
if (server.daemonize) createPidFile();
redisSetProcTitle(argv[0]);
redisAsciiArt();
checkTcpBacklogSettings();
if (!server.sentinel_mode) {
/* Things not needed when running in Sentinel mode. */
@@ -3658,7 +3668,6 @@ int main(int argc, char **argv) {
#ifdef __linux__
linuxMemoryWarnings();
#endif
checkTcpBacklogSettings();
loadDataFromDisk();
if (server.cluster_enabled) {
if (verifyClusterConfigWithData() == REDIS_ERR) {
src/redis.h
+27 -8
View File
@@ -267,22 +267,35 @@ typedef long long mstime_t; /* millisecond time type. */
#define REDIS_CLIENT_TYPE_PUBSUB 2 /* Clients subscribed to PubSub channels. */
#define REDIS_CLIENT_TYPE_COUNT 3
/* Slave replication state - from the point of view of the slave. */
/* Slave replication state. Used in server.repl_state for slaves to remember
* what to do next. */
#define REDIS_REPL_NONE 0 /* No active replication */
#define REDIS_REPL_CONNECT 1 /* Must connect to master */
#define REDIS_REPL_CONNECTING 2 /* Connecting to master */
#define REDIS_REPL_RECEIVE_PONG 3 /* Wait for PING reply */
#define REDIS_REPL_TRANSFER 4 /* Receiving .rdb from master */
#define REDIS_REPL_CONNECTED 5 /* Connected to master */
#define REDIS_REPL_SEND_AUTH 4 /* Send AUTH to master */
#define REDIS_REPL_RECEIVE_AUTH 5 /* Wait for AUTH reply */
#define REDIS_REPL_SEND_PORT 6 /* Send REPLCONF listening-port */
#define REDIS_REPL_RECEIVE_PORT 7 /* Wait for REPLCONF reply */
#define REDIS_REPL_SEND_CAPA 8 /* Send REPLCONF capa */
#define REDIS_REPL_RECEIVE_CAPA 9 /* Wait for REPLCONF reply */
#define REDIS_REPL_SEND_PSYNC 10 /* Send PSYNC */
#define REDIS_REPL_RECEIVE_PSYNC 11 /* Wait for PSYNC reply */
#define REDIS_REPL_TRANSFER 12 /* Receiving .rdb from master */
#define REDIS_REPL_CONNECTED 13 /* Connected to master */
/* Slave replication state - from the point of view of the master.
/* State of slaves from the POV of the master. Used in client->replstate.
* In SEND_BULK and ONLINE state the slave receives new updates
* in its output queue. In the WAIT_BGSAVE state instead the server is waiting
* to start the next background saving in order to send updates to it. */
#define REDIS_REPL_WAIT_BGSAVE_START 6 /* We need to produce a new RDB file. */
#define REDIS_REPL_WAIT_BGSAVE_END 7 /* Waiting RDB file creation to finish. */
#define REDIS_REPL_SEND_BULK 8 /* Sending RDB file to slave. */
#define REDIS_REPL_ONLINE 9 /* RDB file transmitted, sending just updates. */
#define REDIS_REPL_WAIT_BGSAVE_START 14 /* We need to produce a new RDB file. */
#define REDIS_REPL_WAIT_BGSAVE_END 15 /* Waiting RDB file creation to finish. */
#define REDIS_REPL_SEND_BULK 16 /* Sending RDB file to slave. */
#define REDIS_REPL_ONLINE 17 /* RDB file transmitted, sending just updates. */
/* Slave capabilities. */
#define SLAVE_CAPA_NONE 0
#define SLAVE_CAPA_EOF (1<<0) /* Can parse the RDB EOF streaming format. */
/* Synchronous read timeout - slave side */
#define REDIS_REPL_SYNCIO_TIMEOUT 5
@@ -543,8 +556,12 @@ typedef struct redisClient {
long long reploff; /* replication offset if this is our master */
long long repl_ack_off; /* replication ack offset, if this is a slave */
long long repl_ack_time;/* replication ack time, if this is a slave */
long long psync_initial_offset; /* FULLRESYNC reply offset other slaves
copying this slave output buffer
should use. */
char replrunid[REDIS_RUN_ID_SIZE+1]; /* master run id if this is a master */
int slave_listening_port; /* As configured with: SLAVECONF listening-port */
int slave_capa; /* Slave capabilities: SLAVE_CAPA_* bitwise OR. */
multiState mstate; /* MULTI/EXEC state */
int btype; /* Type of blocking op if REDIS_BLOCKED. */
blockingState bpop; /* blocking state */
@@ -1171,6 +1188,8 @@ int replicationCountAcksByOffset(long long offset);
void replicationSendNewlineToMaster(void);
long long replicationGetSlaveOffset(void);
char *replicationGetSlaveName(redisClient *c);
long long getPsyncInitialOffset(void);
int replicationSetupSlaveForFullResync(redisClient *slave, long long offset);
/* Generic persistence functions */
void startLoading(FILE *fp);
src/replication.c
+408 -166
View File
@@ -201,6 +201,7 @@ void replicationFeedSlaves(list *slaves, int dictid, robj **argv, int argc) {
listRewind(slaves,&li);
while((ln = listNext(&li))) {
redisClient *slave = ln->value;
if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_START) continue;
addReply(slave,selectcmd);
}
@@ -349,6 +350,58 @@ long long addReplyReplicationBacklog(redisClient *c, long long offset) {
return server.repl_backlog_histlen - skip;
}
/* Return the offset to provide as reply to the PSYNC command received
* from the slave. The returned value is only valid immediately after
* the BGSAVE process started and before executing any other command
* from clients. */
long long getPsyncInitialOffset(void) {
long long psync_offset = server.master_repl_offset;
/* Add 1 to psync_offset if it the replication backlog does not exists
* as when it will be created later we'll increment the offset by one. */
if (server.repl_backlog == NULL) psync_offset++;
return psync_offset;
}
/* Send a FULLRESYNC reply in the specific case of a full resynchronization,
* as a side effect setup the slave for a full sync in different ways:
*
* 1) Remember, into the slave client structure, the offset we sent
* here, so that if new slaves will later attach to the same
* background RDB saving process (by duplicating this client output
* buffer), we can get the right offset from this slave.
* 2) Set the replication state of the slave to WAIT_BGSAVE_END so that
* we start accumulating differences from this point.
* 3) Force the replication stream to re-emit a SELECT statement so
* the new slave incremental differences will start selecting the
* right database number.
*
* Normally this function should be called immediately after a successful
* BGSAVE for replication was started, or when there is one already in
* progress that we attached our slave to. */
int replicationSetupSlaveForFullResync(redisClient *slave, long long offset) {
char buf[128];
int buflen;
slave->psync_initial_offset = offset;
slave->replstate = REDIS_REPL_WAIT_BGSAVE_END;
/* We are going to accumulate the incremental changes for this
* slave as well. Set slaveseldb to -1 in order to force to re-emit
* a SLEECT statement in the replication stream. */
server.slaveseldb = -1;
/* Don't send this reply to slaves that approached us with
* the old SYNC command. */
if (!(slave->flags & REDIS_PRE_PSYNC)) {
buflen = snprintf(buf,sizeof(buf),"+FULLRESYNC %s %lld\r\n",
server.runid,offset);
if (write(slave->fd,buf,buflen) != buflen) {
freeClientAsync(slave);
return REDIS_ERR;
}
}
return REDIS_OK;
}
/* This function handles the PSYNC command from the point of view of a
* master receiving a request for partial resynchronization.
*
@@ -422,18 +475,10 @@ int masterTryPartialResynchronization(redisClient *c) {
return REDIS_OK; /* The caller can return, no full resync needed. */
need_full_resync:
/* We need a full resync for some reason... notify the client. */
psync_offset = server.master_repl_offset;
/* Add 1 to psync_offset if it the replication backlog does not exists
* as when it will be created later we'll increment the offset by one. */
if (server.repl_backlog == NULL) psync_offset++;
/* Again, we can't use the connection buffers (see above). */
buflen = snprintf(buf,sizeof(buf),"+FULLRESYNC %s %lld\r\n",
server.runid,psync_offset);
if (write(c->fd,buf,buflen) != buflen) {
freeClientAsync(c);
return REDIS_OK;
}
/* We need a full resync for some reason... Note that we can't
* reply to PSYNC right now if a full SYNC is needed. The reply
* must include the master offset at the time the RDB file we transfer
* is generated, so we need to delay the reply to that moment. */
return REDIS_ERR;
}
@@ -441,18 +486,68 @@ need_full_resync:
* socket target depending on the configuration, and making sure that
* the script cache is flushed before to start.
*
* The mincapa argument is the bitwise AND among all the slaves capabilities
* of the slaves waiting for this BGSAVE, so represents the slave capabilities
* all the slaves support. Can be tested via SLAVE_CAPA_* macros.
*
* Side effects, other than starting a BGSAVE:
*
* 1) Handle the slaves in WAIT_START state, by preparing them for a full
* sync if the BGSAVE was succesfully started, or sending them an error
* and dropping them from the list of slaves.
*
* 2) Flush the Lua scripting script cache if the BGSAVE was actually
* started.
*
* Returns REDIS_OK on success or REDIS_ERR otherwise. */
int startBgsaveForReplication(void) {
int startBgsaveForReplication(int mincapa) {
int retval;
int socket_target = server.repl_diskless_sync && (mincapa & SLAVE_CAPA_EOF);
listIter li;
listNode *ln;
redisLog(REDIS_NOTICE,"Starting BGSAVE for SYNC with target: %s",
server.repl_diskless_sync ? "slaves sockets" : "disk");
socket_target ? "slaves sockets" : "disk");
if (server.repl_diskless_sync)
if (socket_target)
retval = rdbSaveToSlavesSockets();
else
retval = rdbSaveBackground(server.rdb_filename);
/* If we failed to BGSAVE, remove the slaves waiting for a full
* resynchorinization from the list of salves, inform them with
* an error about what happened, close the connection ASAP. */
if (retval == REDIS_ERR) {
redisLog(REDIS_WARNING,"BGSAVE for replication failed");
listRewind(server.slaves,&li);
while((ln = listNext(&li))) {
redisClient *slave = ln->value;
if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_START) {
slave->flags &= ~REDIS_SLAVE;
listDelNode(server.slaves,ln);
addReplyError(slave,
"BGSAVE failed, replication can't continue");
slave->flags |= REDIS_CLOSE_AFTER_REPLY;
}
}
return retval;
}
/* If the target is socket, rdbSaveToSlavesSockets() already setup
* the salves for a full resync. Otherwise for disk target do it now.*/
if (!socket_target) {
listRewind(server.slaves,&li);
while((ln = listNext(&li))) {
redisClient *slave = ln->value;
if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_START) {
replicationSetupSlaveForFullResync(slave,
getPsyncInitialOffset());
}
}
}
/* Flush the script cache, since we need that slave differences are
* accumulated without requiring slaves to match our cached scripts. */
if (retval == REDIS_OK) replicationScriptCacheFlush();
@@ -515,8 +610,16 @@ void syncCommand(redisClient *c) {
/* Full resynchronization. */
server.stat_sync_full++;
/* Here we need to check if there is a background saving operation
* in progress, or if it is required to start one */
/* Setup the slave as one waiting for BGSAVE to start. The following code
* paths will change the state if we handle the slave differently. */
c->replstate = REDIS_REPL_WAIT_BGSAVE_START;
if (server.repl_disable_tcp_nodelay)
anetDisableTcpNoDelay(NULL, c->fd); /* Non critical if it fails. */
c->repldbfd = -1;
c->flags |= REDIS_SLAVE;
listAddNodeTail(server.slaves,c);
/* CASE 1: BGSAVE is in progress, with disk target. */
if (server.rdb_child_pid != -1 &&
server.rdb_child_type == REDIS_RDB_CHILD_TYPE_DISK)
{
@@ -532,51 +635,45 @@ void syncCommand(redisClient *c) {
slave = ln->value;
if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_END) break;
}
if (ln) {
/* To attach this slave, we check that it has at least all the
* capabilities of the slave that triggered the current BGSAVE. */
if (ln && ((c->slave_capa & slave->slave_capa) == slave->slave_capa)) {
/* Perfect, the server is already registering differences for
* another slave. Set the right state, and copy the buffer. */
copyClientOutputBuffer(c,slave);
c->replstate = REDIS_REPL_WAIT_BGSAVE_END;
replicationSetupSlaveForFullResync(c,slave->psync_initial_offset);
redisLog(REDIS_NOTICE,"Waiting for end of BGSAVE for SYNC");
} else {
/* No way, we need to wait for the next BGSAVE in order to
* register differences. */
c->replstate = REDIS_REPL_WAIT_BGSAVE_START;
redisLog(REDIS_NOTICE,"Waiting for next BGSAVE for SYNC");
}
/* CASE 2: BGSAVE is in progress, with socket target. */
} else if (server.rdb_child_pid != -1 &&
server.rdb_child_type == REDIS_RDB_CHILD_TYPE_SOCKET)
{
/* There is an RDB child process but it is writing directly to
* children sockets. We need to wait for the next BGSAVE
* in order to synchronize. */
c->replstate = REDIS_REPL_WAIT_BGSAVE_START;
redisLog(REDIS_NOTICE,"Waiting for next BGSAVE for SYNC");
/* CASE 3: There is no BGSAVE is progress. */
} else {
if (server.repl_diskless_sync) {
if (server.repl_diskless_sync && (c->slave_capa & SLAVE_CAPA_EOF)) {
/* Diskless replication RDB child is created inside
* replicationCron() since we want to delay its start a
* few seconds to wait for more slaves to arrive. */
c->replstate = REDIS_REPL_WAIT_BGSAVE_START;
if (server.repl_diskless_sync_delay)
redisLog(REDIS_NOTICE,"Delay next BGSAVE for SYNC");
} else {
/* Ok we don't have a BGSAVE in progress, let's start one. */
if (startBgsaveForReplication() != REDIS_OK) {
redisLog(REDIS_NOTICE,"Replication failed, can't BGSAVE");
addReplyError(c,"Unable to perform background save");
return;
}
c->replstate = REDIS_REPL_WAIT_BGSAVE_END;
/* Target is disk (or the slave is not capable of supporting
* diskless replication) and we don't have a BGSAVE in progress,
* let's start one. */
if (startBgsaveForReplication(c->slave_capa) != REDIS_OK) return;
}
}
if (server.repl_disable_tcp_nodelay)
anetDisableTcpNoDelay(NULL, c->fd); /* Non critical if it fails. */
c->repldbfd = -1;
c->flags |= REDIS_SLAVE;
server.slaveseldb = -1; /* Force to re-emit the SELECT command. */
listAddNodeTail(server.slaves,c);
if (listLength(server.slaves) == 1 && server.repl_backlog == NULL)
createReplicationBacklog();
return;
@@ -613,6 +710,10 @@ void replconfCommand(redisClient *c) {
&port,NULL) != REDIS_OK))
return;
c->slave_listening_port = port;
} else if (!strcasecmp(c->argv[j]->ptr,"capa")) {
/* Ignore capabilities not understood by this master. */
if (!strcasecmp(c->argv[j+1]->ptr,"eof"))
c->slave_capa |= SLAVE_CAPA_EOF;
} else if (!strcasecmp(c->argv[j]->ptr,"ack")) {
/* REPLCONF ACK is used by slave to inform the master the amount
* of replication stream that it processed so far. It is an
@@ -746,6 +847,7 @@ void sendBulkToSlave(aeEventLoop *el, int fd, void *privdata, int mask) {
void updateSlavesWaitingBgsave(int bgsaveerr, int type) {
listNode *ln;
int startbgsave = 0;
int mincapa = -1;
listIter li;
listRewind(server.slaves,&li);
@@ -754,7 +856,8 @@ void updateSlavesWaitingBgsave(int bgsaveerr, int type) {
if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_START) {
startbgsave = 1;
slave->replstate = REDIS_REPL_WAIT_BGSAVE_END;
mincapa = (mincapa == -1) ? slave->slave_capa :
(mincapa & slave->slave_capa);
} else if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_END) {
struct redis_stat buf;
@@ -801,20 +904,7 @@ void updateSlavesWaitingBgsave(int bgsaveerr, int type) {
}
}
}
if (startbgsave) {
if (startBgsaveForReplication() != REDIS_OK) {
listIter li;
listRewind(server.slaves,&li);
redisLog(REDIS_WARNING,"SYNC failed. BGSAVE failed");
while((ln = listNext(&li))) {
redisClient *slave = ln->value;
if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_START)
freeClient(slave);
}
}
}
if (startbgsave) startBgsaveForReplication(mincapa);
}
/* ----------------------------------- SLAVE -------------------------------- */
@@ -1062,38 +1152,53 @@ error:
* The command returns an sds string representing the result of the
* operation. On error the first byte is a "-".
*/
char *sendSynchronousCommand(int fd, ...) {
va_list ap;
sds cmd = sdsempty();
char *arg, buf[256];
#define SYNC_CMD_READ (1<<0)
#define SYNC_CMD_WRITE (1<<1)
#define SYNC_CMD_FULL (SYNC_CMD_READ|SYNC_CMD_WRITE)
char *sendSynchronousCommand(int flags, int fd, ...) {
/* Create the command to send to the master, we use simple inline
* protocol for simplicity as currently we only send simple strings. */
va_start(ap,fd);
while(1) {
arg = va_arg(ap, char*);
if (arg == NULL) break;
if (flags & SYNC_CMD_WRITE) {
char *arg;
va_list ap;
sds cmd = sdsempty();
va_start(ap,fd);
if (sdslen(cmd) != 0) cmd = sdscatlen(cmd," ",1);
cmd = sdscat(cmd,arg);
}
cmd = sdscatlen(cmd,"\r\n",2);
while(1) {
arg = va_arg(ap, char*);
if (arg == NULL) break;
/* Transfer command to the server. */
if (syncWrite(fd,cmd,sdslen(cmd),server.repl_syncio_timeout*1000) == -1) {
if (sdslen(cmd) != 0) cmd = sdscatlen(cmd," ",1);
cmd = sdscat(cmd,arg);
}
cmd = sdscatlen(cmd,"\r\n",2);
/* Transfer command to the server. */
if (syncWrite(fd,cmd,sdslen(cmd),server.repl_syncio_timeout*1000)
== -1)
{
sdsfree(cmd);
return sdscatprintf(sdsempty(),"-Writing to master: %s",
strerror(errno));
}
sdsfree(cmd);
return sdscatprintf(sdsempty(),"-Writing to master: %s",
strerror(errno));
va_end(ap);
}
sdsfree(cmd);
/* Read the reply from the server. */
if (syncReadLine(fd,buf,sizeof(buf),server.repl_syncio_timeout*1000) == -1)
{
return sdscatprintf(sdsempty(),"-Reading from master: %s",
strerror(errno));
if (flags & SYNC_CMD_READ) {
char buf[256];
if (syncReadLine(fd,buf,sizeof(buf),server.repl_syncio_timeout*1000)
== -1)
{
return sdscatprintf(sdsempty(),"-Reading from master: %s",
strerror(errno));
}
return sdsnew(buf);
}
return sdsnew(buf);
return NULL;
}
/* Try a partial resynchronization with the master if we are about to reconnect.
@@ -1110,6 +1215,19 @@ char *sendSynchronousCommand(int fd, ...) {
* of successful partial resynchronization, the function will reuse
* 'fd' as file descriptor of the server.master client structure.
*
* The function is split in two halves: if read_reply is 0, the function
* writes the PSYNC command on the socket, and a new function call is
* needed, with read_reply set to 1, in order to read the reply of the
* command. This is useful in order to support non blocking operations, so
* that we write, return into the event loop, and read when there are data.
*
* When read_reply is 0 the function returns PSYNC_WRITE_ERR if there
* was a write error, or PSYNC_WAIT_REPLY to signal we need another call
* with read_reply set to 1. However even when read_reply is set to 1
* the function may return PSYNC_WAIT_REPLY again to signal there were
* insufficient data to read to complete its work. We should re-enter
* into the event loop and wait in such a case.
*
* The function returns:
*
* PSYNC_CONTINUE: If the PSYNC command succeded and we can continue.
@@ -1118,35 +1236,68 @@ char *sendSynchronousCommand(int fd, ...) {
* offset is saved.
* PSYNC_NOT_SUPPORTED: If the server does not understand PSYNC at all and
* the caller should fall back to SYNC.
* PSYNC_WRITE_ERR: There was an error writing the command to the socket.
* PSYNC_WAIT_REPLY: Call again the function with read_reply set to 1.
*
* Notable side effects:
*
* 1) As a side effect of the function call the function removes the readable
* event handler from "fd", unless the return value is PSYNC_WAIT_REPLY.
* 2) server.repl_master_initial_offset is set to the right value according
* to the master reply. This will be used to populate the 'server.master'
* structure replication offset.
*/
#define PSYNC_CONTINUE 0
#define PSYNC_FULLRESYNC 1
#define PSYNC_NOT_SUPPORTED 2
int slaveTryPartialResynchronization(int fd) {
#define PSYNC_WRITE_ERROR 0
#define PSYNC_WAIT_REPLY 1
#define PSYNC_CONTINUE 2
#define PSYNC_FULLRESYNC 3
#define PSYNC_NOT_SUPPORTED 4
int slaveTryPartialResynchronization(int fd, int read_reply) {
char *psync_runid;
char psync_offset[32];
sds reply;
/* Initially set repl_master_initial_offset to -1 to mark the current
* master run_id and offset as not valid. Later if we'll be able to do
* a FULL resync using the PSYNC command we'll set the offset at the
* right value, so that this information will be propagated to the
* client structure representing the master into server.master. */
server.repl_master_initial_offset = -1;
/* Writing half */
if (!read_reply) {
/* Initially set repl_master_initial_offset to -1 to mark the current
* master run_id and offset as not valid. Later if we'll be able to do
* a FULL resync using the PSYNC command we'll set the offset at the
* right value, so that this information will be propagated to the
* client structure representing the master into server.master. */
server.repl_master_initial_offset = -1;
if (server.cached_master) {
psync_runid = server.cached_master->replrunid;
snprintf(psync_offset,sizeof(psync_offset),"%lld", server.cached_master->reploff+1);
redisLog(REDIS_NOTICE,"Trying a partial resynchronization (request %s:%s).", psync_runid, psync_offset);
} else {
redisLog(REDIS_NOTICE,"Partial resynchronization not possible (no cached master)");
psync_runid = "?";
memcpy(psync_offset,"-1",3);
if (server.cached_master) {
psync_runid = server.cached_master->replrunid;
snprintf(psync_offset,sizeof(psync_offset),"%lld", server.cached_master->reploff+1);
redisLog(REDIS_NOTICE,"Trying a partial resynchronization (request %s:%s).", psync_runid, psync_offset);
} else {
redisLog(REDIS_NOTICE,"Partial resynchronization not possible (no cached master)");
psync_runid = "?";
memcpy(psync_offset,"-1",3);
}
/* Issue the PSYNC command */
reply = sendSynchronousCommand(SYNC_CMD_WRITE,fd,"PSYNC",psync_runid,psync_offset,NULL);
if (reply != NULL) {
redisLog(REDIS_WARNING,"Unable to send PSYNC to master: %s",reply);
sdsfree(reply);
aeDeleteFileEvent(server.el,fd,AE_READABLE);
return PSYNC_WRITE_ERROR;
}
return PSYNC_WAIT_REPLY;
}
/* Issue the PSYNC command */
reply = sendSynchronousCommand(fd,"PSYNC",psync_runid,psync_offset,NULL);
/* Reading half */
reply = sendSynchronousCommand(SYNC_CMD_READ,fd,NULL);
if (sdslen(reply) == 0) {
/* The master may send empty newlines after it receives PSYNC
* and before to reply, just to keep the connection alive. */
sdsfree(reply);
return PSYNC_WAIT_REPLY;
}
aeDeleteFileEvent(server.el,fd,AE_READABLE);
if (!strncmp(reply,"+FULLRESYNC",11)) {
char *runid = NULL, *offset = NULL;
@@ -1190,7 +1341,7 @@ int slaveTryPartialResynchronization(int fd) {
return PSYNC_CONTINUE;
}
/* If we reach this point we receied either an error since the master does
/* If we reach this point we received either an error since the master does
* not understand PSYNC, or an unexpected reply from the master.
* Return PSYNC_NOT_SUPPORTED to the caller in both cases. */
@@ -1209,7 +1360,7 @@ int slaveTryPartialResynchronization(int fd) {
}
void syncWithMaster(aeEventLoop *el, int fd, void *privdata, int mask) {
char tmpfile[256], *err;
char tmpfile[256], *err = NULL;
int dfd, maxtries = 5;
int sockerr = 0, psync_result;
socklen_t errlen = sizeof(sockerr);
@@ -1228,16 +1379,12 @@ void syncWithMaster(aeEventLoop *el, int fd, void *privdata, int mask) {
if (getsockopt(fd, SOL_SOCKET, SO_ERROR, &sockerr, &errlen) == -1)
sockerr = errno;
if (sockerr) {
aeDeleteFileEvent(server.el,fd,AE_READABLE|AE_WRITABLE);
redisLog(REDIS_WARNING,"Error condition on socket for SYNC: %s",
strerror(sockerr));
goto error;
}
/* If we were connecting, it's time to send a non blocking PING, we want to
* make sure the master is able to reply before going into the actual
* replication process where we have long timeouts in the order of
* seconds (in the meantime the slave would block). */
/* Send a PING to check the master is able to reply without errors. */
if (server.repl_state == REDIS_REPL_CONNECTING) {
redisLog(REDIS_NOTICE,"Non blocking connect for SYNC fired the event.");
/* Delete the writable event so that the readable event remains
@@ -1246,70 +1393,109 @@ void syncWithMaster(aeEventLoop *el, int fd, void *privdata, int mask) {
server.repl_state = REDIS_REPL_RECEIVE_PONG;
/* Send the PING, don't check for errors at all, we have the timeout
* that will take care about this. */
syncWrite(fd,"PING\r\n",6,100);
err = sendSynchronousCommand(SYNC_CMD_WRITE,fd,"PING",NULL);
if (err) goto write_error;
return;
}
/* Receive the PONG command. */
if (server.repl_state == REDIS_REPL_RECEIVE_PONG) {
char buf[1024];
/* Delete the readable event, we no longer need it now that there is
* the PING reply to read. */
aeDeleteFileEvent(server.el,fd,AE_READABLE);
/* Read the reply with explicit timeout. */
buf[0] = '\0';
if (syncReadLine(fd,buf,sizeof(buf),
server.repl_syncio_timeout*1000) == -1)
{
redisLog(REDIS_WARNING,
"I/O error reading PING reply from master: %s",
strerror(errno));
goto error;
}
err = sendSynchronousCommand(SYNC_CMD_READ,fd,NULL);
/* We accept only two replies as valid, a positive +PONG reply
* (we just check for "+") or an authentication error.
* Note that older versions of Redis replied with "operation not
* permitted" instead of using a proper error code, so we test
* both. */
if (buf[0] != '+' &&
strncmp(buf,"-NOAUTH",7) != 0 &&
strncmp(buf,"-ERR operation not permitted",28) != 0)
if (err[0] != '+' &&
strncmp(err,"-NOAUTH",7) != 0 &&
strncmp(err,"-ERR operation not permitted",28) != 0)
{
redisLog(REDIS_WARNING,"Error reply to PING from master: '%s'",buf);
redisLog(REDIS_WARNING,"Error reply to PING from master: '%s'",err);
sdsfree(err);
goto error;
} else {
redisLog(REDIS_NOTICE,
"Master replied to PING, replication can continue...");
}
sdsfree(err);
server.repl_state = REDIS_REPL_SEND_AUTH;
}
/* AUTH with the master if required. */
if(server.masterauth) {
err = sendSynchronousCommand(fd,"AUTH",server.masterauth,NULL);
if (server.repl_state == REDIS_REPL_SEND_AUTH) {
if (server.masterauth) {
err = sendSynchronousCommand(SYNC_CMD_WRITE,fd,"AUTH",server.masterauth,NULL);
if (err) goto write_error;
server.repl_state = REDIS_REPL_RECEIVE_AUTH;
return;
} else {
server.repl_state = REDIS_REPL_SEND_PORT;
}
}
/* Receive AUTH reply. */
if (server.repl_state == REDIS_REPL_RECEIVE_AUTH) {
err = sendSynchronousCommand(SYNC_CMD_READ,fd,NULL);
if (err[0] == '-') {
redisLog(REDIS_WARNING,"Unable to AUTH to MASTER: %s",err);
sdsfree(err);
goto error;
}
sdsfree(err);
server.repl_state = REDIS_REPL_SEND_PORT;
}
/* Set the slave port, so that Master's INFO command can list the
* slave listening port correctly. */
{
if (server.repl_state == REDIS_REPL_SEND_PORT) {
sds port = sdsfromlonglong(server.port);
err = sendSynchronousCommand(fd,"REPLCONF","listening-port",port,
NULL);
err = sendSynchronousCommand(SYNC_CMD_WRITE,fd,"REPLCONF",
"listening-port",port, NULL);
sdsfree(port);
if (err) goto write_error;
sdsfree(err);
server.repl_state = REDIS_REPL_RECEIVE_PORT;
return;
}
/* Receive REPLCONF listening-port reply. */
if (server.repl_state == REDIS_REPL_RECEIVE_PORT) {
err = sendSynchronousCommand(SYNC_CMD_READ,fd,NULL);
/* Ignore the error if any, not all the Redis versions support
* REPLCONF listening-port. */
if (err[0] == '-') {
redisLog(REDIS_NOTICE,"(Non critical) Master does not understand REPLCONF listening-port: %s", err);
redisLog(REDIS_NOTICE,"(Non critical) Master does not understand "
"REPLCONF listening-port: %s", err);
}
sdsfree(err);
server.repl_state = REDIS_REPL_SEND_CAPA;
}
/* Inform the master of our capabilities. While we currently send
* just one capability, it is possible to chain new capabilities here
* in the form of REPLCONF capa X capa Y capa Z ...
* The master will ignore capabilities it does not understand. */
if (server.repl_state == REDIS_REPL_SEND_CAPA) {
err = sendSynchronousCommand(SYNC_CMD_WRITE,fd,"REPLCONF",
"capa","eof",NULL);
if (err) goto write_error;
sdsfree(err);
server.repl_state = REDIS_REPL_RECEIVE_CAPA;
return;
}
/* Receive CAPA reply. */
if (server.repl_state == REDIS_REPL_RECEIVE_CAPA) {
err = sendSynchronousCommand(SYNC_CMD_READ,fd,NULL);
/* Ignore the error if any, not all the Redis versions support
* REPLCONF capa. */
if (err[0] == '-') {
redisLog(REDIS_NOTICE,"(Non critical) Master does not understand "
"REPLCONF capa: %s", err);
}
sdsfree(err);
server.repl_state = REDIS_REPL_SEND_PSYNC;
}
/* Try a partial resynchonization. If we don't have a cached master
@@ -1317,12 +1503,41 @@ void syncWithMaster(aeEventLoop *el, int fd, void *privdata, int mask) {
* to start a full resynchronization so that we get the master run id
* and the global offset, to try a partial resync at the next
* reconnection attempt. */
psync_result = slaveTryPartialResynchronization(fd);
if (server.repl_state == REDIS_REPL_SEND_PSYNC) {
if (slaveTryPartialResynchronization(fd,0) == PSYNC_WRITE_ERROR) {
err = sdsnew("Write error sending the PSYNC command.");
goto write_error;
}
server.repl_state = REDIS_REPL_RECEIVE_PSYNC;
return;
}
/* If reached this point, we should be in REDIS_REPL_RECEIVE_PSYNC. */
if (server.repl_state != REDIS_REPL_RECEIVE_PSYNC) {
redisLog(REDIS_WARNING,"syncWithMaster(): state machine error, "
"state should be RECEIVE_PSYNC but is %d",
server.repl_state);
goto error;
}
psync_result = slaveTryPartialResynchronization(fd,1);
if (psync_result == PSYNC_WAIT_REPLY) return; /* Try again later... */
/* Note: if PSYNC does not return WAIT_REPLY, it will take care of
* uninstalling the read handler from the file descriptor. */
if (psync_result == PSYNC_CONTINUE) {
redisLog(REDIS_NOTICE, "MASTER <-> SLAVE sync: Master accepted a Partial Resynchronization.");
return;
}
/* PSYNC failed or is not supported: we want our slaves to resync with us
* as well, if we have any (chained replication case). The mater may
* transfer us an entirely different data set and we have no way to
* incrementally feed our slaves after that. */
disconnectSlaves(); /* Force our slaves to resync with us as well. */
freeReplicationBacklog(); /* Don't allow our chained slaves to PSYNC. */
/* Fall back to SYNC if needed. Otherwise psync_result == PSYNC_FULLRESYNC
* and the server.repl_master_runid and repl_master_initial_offset are
* already populated. */
@@ -1368,16 +1583,22 @@ void syncWithMaster(aeEventLoop *el, int fd, void *privdata, int mask) {
return;
error:
aeDeleteFileEvent(server.el,fd,AE_READABLE|AE_WRITABLE);
close(fd);
server.repl_transfer_s = -1;
server.repl_state = REDIS_REPL_CONNECT;
return;
write_error: /* Handle sendSynchronousCommand(SYNC_CMD_WRITE) errors. */
redisLog(REDIS_WARNING,"Sending command to master in replication handshake: %s", err);
sdsfree(err);
goto error;
}
int connectWithMaster(void) {
int fd;
fd = anetTcpNonBlockBindConnect(NULL,
fd = anetTcpNonBlockBestEffortBindConnect(NULL,
server.masterhost,server.masterport,REDIS_BIND_ADDR);
if (fd == -1) {
redisLog(REDIS_WARNING,"Unable to connect to MASTER: %s",
@@ -1470,6 +1691,17 @@ void replicationUnsetMaster(void) {
server.repl_state = REDIS_REPL_NONE;
}
/* This function is called when the slave lose the connection with the
* master into an unexpected way. */
void replicationHandleMasterDisconnection(void) {
server.master = NULL;
server.repl_state = REDIS_REPL_CONNECT;
server.repl_down_since = server.unixtime;
/* We lost connection with our master, don't disconnect slaves yet,
* maybe we'll be able to PSYNC with our master later. We'll disconnect
* the slaves only if we'll have to do a full resync with our master. */
}
void slaveofCommand(redisClient *c) {
/* SLAVEOF is not allowed in cluster mode as replication is automatically
* configured using the current address of the master node. */
@@ -1484,7 +1716,10 @@ void slaveofCommand(redisClient *c) {
!strcasecmp(c->argv[2]->ptr,"one")) {
if (server.masterhost) {
replicationUnsetMaster();
redisLog(REDIS_NOTICE,"MASTER MODE enabled (user request)");
sds client = catClientInfoString(sdsempty(),c);
redisLog(REDIS_NOTICE,
"MASTER MODE enabled (user request from '%s')",client);
sdsfree(client);
}
} else {
long port;
@@ -1502,8 +1737,10 @@ void slaveofCommand(redisClient *c) {
/* There was no previous master or the user specified a different one,
* we can continue. */
replicationSetMaster(c->argv[1]->ptr, port);
redisLog(REDIS_NOTICE,"SLAVE OF %s:%d enabled (user request)",
server.masterhost, server.masterport);
sds client = catClientInfoString(sdsempty(),c);
redisLog(REDIS_NOTICE,"SLAVE OF %s:%d enabled (user request from '%s')",
server.masterhost, server.masterport, client);
sdsfree(client);
}
addReply(c,shared.ok);
}
@@ -1547,7 +1784,15 @@ void roleCommand(redisClient *c) {
case REDIS_REPL_NONE: slavestate = "none"; break;
case REDIS_REPL_CONNECT: slavestate = "connect"; break;
case REDIS_REPL_CONNECTING: slavestate = "connecting"; break;
case REDIS_REPL_RECEIVE_PONG: /* see next */
case REDIS_REPL_RECEIVE_PONG:
case REDIS_REPL_SEND_AUTH:
case REDIS_REPL_RECEIVE_AUTH:
case REDIS_REPL_SEND_PORT:
case REDIS_REPL_RECEIVE_PORT:
case REDIS_REPL_SEND_CAPA:
case REDIS_REPL_RECEIVE_CAPA:
case REDIS_REPL_SEND_PSYNC:
case REDIS_REPL_RECEIVE_PSYNC: slavestate = "handshake"; break;
case REDIS_REPL_TRANSFER: slavestate = "sync"; break;
case REDIS_REPL_CONNECTED: slavestate = "connected"; break;
default: slavestate = "unknown"; break;
@@ -1936,6 +2181,8 @@ long long replicationGetSlaveOffset(void) {
/* Replication cron function, called 1 time per second. */
void replicationCron(void) {
static long long replication_cron_loops = 0;
/* Non blocking connection timeout? */
if (server.masterhost &&
(server.repl_state == REDIS_REPL_CONNECTING ||
@@ -1982,31 +2229,34 @@ void replicationCron(void) {
* So slaves can implement an explicit timeout to masters, and will
* be able to detect a link disconnection even if the TCP connection
* will not actually go down. */
if (!(server.cronloops % (server.repl_ping_slave_period * server.hz))) {
listIter li;
listNode *ln;
robj *ping_argv[1];
listIter li;
listNode *ln;
robj *ping_argv[1];
/* First, send PING */
/* First, send PING according to ping_slave_period. */
if ((replication_cron_loops % server.repl_ping_slave_period) == 0) {
ping_argv[0] = createStringObject("PING",4);
replicationFeedSlaves(server.slaves, server.slaveseldb, ping_argv, 1);
replicationFeedSlaves(server.slaves, server.slaveseldb,
ping_argv, 1);
decrRefCount(ping_argv[0]);
}
/* Second, send a newline to all the slaves in pre-synchronization
* stage, that is, slaves waiting for the master to create the RDB file.
* The newline will be ignored by the slave but will refresh the
* last-io timer preventing a timeout. */
listRewind(server.slaves,&li);
while((ln = listNext(&li))) {
redisClient *slave = ln->value;
/* Second, send a newline to all the slaves in pre-synchronization
* stage, that is, slaves waiting for the master to create the RDB file.
* The newline will be ignored by the slave but will refresh the
* last-io timer preventing a timeout. In this case we ignore the
* ping period and refresh the connection once per second since certain
* timeouts are set at a few seconds (example: PSYNC response). */
listRewind(server.slaves,&li);
while((ln = listNext(&li))) {
redisClient *slave = ln->value;
if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_START ||
(slave->replstate == REDIS_REPL_WAIT_BGSAVE_END &&
server.rdb_child_type != REDIS_RDB_CHILD_TYPE_SOCKET))
{
if (write(slave->fd, "\n", 1) == -1) {
/* Don't worry, it's just a ping. */
}
if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_START ||
(slave->replstate == REDIS_REPL_WAIT_BGSAVE_END &&
server.rdb_child_type != REDIS_RDB_CHILD_TYPE_SOCKET))
{
if (write(slave->fd, "\n", 1) == -1) {
/* Don't worry, it's just a ping. */
}
}
}
@@ -2067,6 +2317,7 @@ void replicationCron(void) {
if (server.rdb_child_pid == -1 && server.aof_child_pid == -1) {
time_t idle, max_idle = 0;
int slaves_waiting = 0;
int mincapa = -1;
listNode *ln;
listIter li;
@@ -2077,28 +2328,19 @@ void replicationCron(void) {
idle = server.unixtime - slave->lastinteraction;
if (idle > max_idle) max_idle = idle;
slaves_waiting++;
mincapa = (mincapa == -1) ? slave->slave_capa :
(mincapa & slave->slave_capa);
}
}
if (slaves_waiting && max_idle > server.repl_diskless_sync_delay) {
/* Start a BGSAVE. Usually with socket target, or with disk target
* if there was a recent socket -> disk config change. */
if (startBgsaveForReplication() == REDIS_OK) {
/* It started! We need to change the state of slaves
* from WAIT_BGSAVE_START to WAIT_BGSAVE_END in case
* the current target is disk. Otherwise it was already done
* by rdbSaveToSlavesSockets() which is called by
* startBgsaveForReplication(). */
listRewind(server.slaves,&li);
while((ln = listNext(&li))) {
redisClient *slave = ln->value;
if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_START)
slave->replstate = REDIS_REPL_WAIT_BGSAVE_END;
}
}
startBgsaveForReplication(mincapa);
}
}
/* Refresh the number of slaves with lag <= min-slaves-max-lag. */
refreshGoodSlavesCount();
replication_cron_loops++; /* Incremented with frequency 1 HZ. */
}
src/scripting.c
+9 -6
View File
@@ -612,11 +612,12 @@ void scriptingEnableGlobalsProtection(lua_State *lua) {
/* strict.lua from: http://metalua.luaforge.net/src/lib/strict.lua.html.
* Modified to be adapted to Redis. */
s[j++]="local dbg=debug\n";
s[j++]="local mt = {}\n";
s[j++]="setmetatable(_G, mt)\n";
s[j++]="mt.__newindex = function (t, n, v)\n";
s[j++]=" if debug.getinfo(2) then\n";
s[j++]=" local w = debug.getinfo(2, \"S\").what\n";
s[j++]=" if dbg.getinfo(2) then\n";
s[j++]=" local w = dbg.getinfo(2, \"S\").what\n";
s[j++]=" if w ~= \"main\" and w ~= \"C\" then\n";
s[j++]=" error(\"Script attempted to create global variable '\"..tostring(n)..\"'\", 2)\n";
s[j++]=" end\n";
@@ -624,11 +625,12 @@ void scriptingEnableGlobalsProtection(lua_State *lua) {
s[j++]=" rawset(t, n, v)\n";
s[j++]="end\n";
s[j++]="mt.__index = function (t, n)\n";
s[j++]=" if debug.getinfo(2) and debug.getinfo(2, \"S\").what ~= \"C\" then\n";
s[j++]=" if dbg.getinfo(2) and dbg.getinfo(2, \"S\").what ~= \"C\" then\n";
s[j++]=" error(\"Script attempted to access unexisting global variable '\"..tostring(n)..\"'\", 2)\n";
s[j++]=" end\n";
s[j++]=" return rawget(t, n)\n";
s[j++]="end\n";
s[j++]="debug = nil\n";
s[j++]=NULL;
for (j = 0; s[j] != NULL; j++) code = sdscatlen(code,s[j],strlen(s[j]));
@@ -732,10 +734,11 @@ void scriptingInit(void) {
* information about the caller, that's what makes sense from the point
* of view of the user debugging a script. */
{
char *errh_func = "function __redis__err__handler(err)\n"
" local i = debug.getinfo(2,'nSl')\n"
char *errh_func = "local dbg = debug\n"
"function __redis__err__handler(err)\n"
" local i = dbg.getinfo(2,'nSl')\n"
" if i and i.what == 'C' then\n"
" i = debug.getinfo(3,'nSl')\n"
" i = dbg.getinfo(3,'nSl')\n"
" end\n"
" if i then\n"
" return i.source .. ':' .. i.currentline .. ': ' .. err\n"
src/sentinel.c
+63 -3
View File
@@ -1157,7 +1157,9 @@ void sentinelDelFlagsToDictOfRedisInstances(dict *instances, int flags) {
* 1) Remove all slaves.
* 2) Remove all sentinels.
* 3) Remove most of the flags resulting from runtime operations.
* 4) Reset timers to their default value.
* 4) Reset timers to their default value. For example after a reset it will be
* possible to failover again the same master ASAP, without waiting the
* failover timeout delay.
* 5) In the process of doing this undo the failover if in progress.
* 6) Disconnect the connections with the master (will reconnect automatically).
*/
@@ -1180,7 +1182,7 @@ void sentinelResetMaster(sentinelRedisInstance *ri, int flags) {
}
ri->failover_state = SENTINEL_FAILOVER_STATE_NONE;
ri->failover_state_change_time = 0;
ri->failover_start_time = 0;
ri->failover_start_time = 0; /* We can failover again ASAP. */
ri->promoted_slave = NULL;
sdsfree(ri->runid);
sdsfree(ri->slave_master_host);
@@ -1544,7 +1546,7 @@ void rewriteConfigSentinelOption(struct rewriteConfigState *state) {
slave_addr = master->addr;
line = sdscatprintf(sdsempty(),
"sentinel known-slave %s %s %d",
master->name, ri->addr->ip, ri->addr->port);
master->name, slave_addr->ip, slave_addr->port);
rewriteConfigRewriteLine(state,"sentinel",line,1);
}
dictReleaseIterator(di2);
@@ -2614,6 +2616,31 @@ sentinelRedisInstance *sentinelGetMasterByNameOrReplyError(redisClient *c,
return ri;
}
#define SENTINEL_ISQR_OK 0
#define SENTINEL_ISQR_NOQUORUM (1<<0)
#define SENTINEL_ISQR_NOAUTH (1<<1)
int sentinelIsQuorumReachable(sentinelRedisInstance *master, int *usableptr) {
dictIterator *di;
dictEntry *de;
int usable = 1; /* Number of usable Sentinels. Init to 1 to count myself. */
int result = SENTINEL_ISQR_OK;
int voters = dictSize(master->sentinels)+1; /* Known Sentinels + myself. */
di = dictGetIterator(master->sentinels);
while((de = dictNext(di)) != NULL) {
sentinelRedisInstance *ri = dictGetVal(de);
if (ri->flags & (SRI_S_DOWN|SRI_O_DOWN)) continue;
usable++;
}
dictReleaseIterator(di);
if (usable < (int)master->quorum) result |= SENTINEL_ISQR_NOQUORUM;
if (usable < voters/2+1) result |= SENTINEL_ISQR_NOAUTH;
if (usableptr) *usableptr = usable;
return result;
}
void sentinelCommand(redisClient *c) {
if (!strcasecmp(c->argv[1]->ptr,"masters")) {
/* SENTINEL MASTERS */
@@ -2764,16 +2791,49 @@ void sentinelCommand(redisClient *c) {
sentinelEvent(REDIS_WARNING,"+monitor",ri,"%@ quorum %d",ri->quorum);
addReply(c,shared.ok);
}
} else if (!strcasecmp(c->argv[1]->ptr,"flushconfig")) {
if (c->argc != 2) goto numargserr;
sentinelFlushConfig();
addReply(c,shared.ok);
return;
} else if (!strcasecmp(c->argv[1]->ptr,"remove")) {
/* SENTINEL REMOVE <name> */
sentinelRedisInstance *ri;
if (c->argc != 3) goto numargserr;
if ((ri = sentinelGetMasterByNameOrReplyError(c,c->argv[2]))
== NULL) return;
sentinelEvent(REDIS_WARNING,"-monitor",ri,"%@");
dictDelete(sentinel.masters,c->argv[2]->ptr);
sentinelFlushConfig();
addReply(c,shared.ok);
} else if (!strcasecmp(c->argv[1]->ptr,"ckquorum")) {
/* SENTINEL CKQUORUM <name> */
sentinelRedisInstance *ri;
int usable;
if (c->argc != 3) goto numargserr;
if ((ri = sentinelGetMasterByNameOrReplyError(c,c->argv[2]))
== NULL) return;
int result = sentinelIsQuorumReachable(ri,&usable);
if (result == SENTINEL_ISQR_OK) {
addReplySds(c, sdscatfmt(sdsempty(),
"+OK %i usable Sentinels. Quorum and failover authorization "
"can be reached\r\n",usable));
} else {
sds e = sdscatfmt(sdsempty(),
"-NOQUORUM %i usable Sentinels. ",usable);
if (result & SENTINEL_ISQR_NOQUORUM)
e = sdscat(e,"Not enough available Sentinels to reach the"
" specified quorum for this master");
if (result & SENTINEL_ISQR_NOAUTH) {
if (result & SENTINEL_ISQR_NOQUORUM) e = sdscat(e,". ");
e = sdscat(e, "Not enough available Sentinels to reach the"
" majority and authorize a failover");
}
e = sdscat(e,"\r\n");
addReplySds(c,e);
}
} else if (!strcasecmp(c->argv[1]->ptr,"set")) {
if (c->argc < 3 || c->argc % 2 == 0) goto numargserr;
sentinelSetCommand(c);
src/sort.c
+1 -1
View File
@@ -209,7 +209,7 @@ void sortCommand(redisClient *c) {
}
/* Create a list of operations to perform for every sorted element.
* Operations can be GET/DEL/INCR/DECR */
* Operations can be GET */
operations = listCreate();
listSetFreeMethod(operations,zfree);
j = 2; /* options start at argv[2] */
src/t_set.c
+1 -1
View File
@@ -321,7 +321,7 @@ void smoveCommand(redisClient *c) {
/* If srcset and dstset are equal, SMOVE is a no-op */
if (srcset == dstset) {
addReply(c,shared.cone);
addReply(c,setTypeIsMember(srcset,ele) ? shared.cone : shared.czero);
return;
}
src/t_zset.c
+78 -16
View File
@@ -1171,35 +1171,84 @@ void zsetConvert(robj *zobj, int encoding) {
*----------------------------------------------------------------------------*/
/* This generic command implements both ZADD and ZINCRBY. */
void zaddGenericCommand(redisClient *c, int incr) {
#define ZADD_NONE 0
#define ZADD_INCR (1<<0) /* Increment the score instead of setting it. */
#define ZADD_NX (1<<1) /* Don't touch elements not already existing. */
#define ZADD_XX (1<<2) /* Only touch elements already exisitng. */
#define ZADD_CH (1<<3) /* Return num of elements added or updated. */
void zaddGenericCommand(redisClient *c, int flags) {
static char *nanerr = "resulting score is not a number (NaN)";
robj *key = c->argv[1];
robj *ele;
robj *zobj;
robj *curobj;
double score = 0, *scores = NULL, curscore = 0.0;
int j, elements = (c->argc-2)/2;
int added = 0, updated = 0;
int j, elements;
int scoreidx = 0;
/* The following vars are used in order to track what the command actually
* did during the execution, to reply to the client and to trigger the
* notification of keyspace change. */
int added = 0; /* Number of new elements added. */
int updated = 0; /* Number of elements with updated score. */
int processed = 0; /* Number of elements processed, may remain zero with
options like XX. */
if (c->argc % 2) {
/* Parse options. At the end 'scoreidx' is set to the argument position
* of the score of the first score-element pair. */
scoreidx = 2;
while(scoreidx < c->argc) {
char *opt = c->argv[scoreidx]->ptr;
if (!strcasecmp(opt,"nx")) flags |= ZADD_NX;
else if (!strcasecmp(opt,"xx")) flags |= ZADD_XX;
else if (!strcasecmp(opt,"ch")) flags |= ZADD_CH;
else if (!strcasecmp(opt,"incr")) flags |= ZADD_INCR;
else break;
scoreidx++;
}
/* Turn options into simple to check vars. */
int incr = (flags & ZADD_INCR) != 0;
int nx = (flags & ZADD_NX) != 0;
int xx = (flags & ZADD_XX) != 0;
int ch = (flags & ZADD_CH) != 0;
/* After the options, we expect to have an even number of args, since
* we expect any number of score-element pairs. */
elements = c->argc-scoreidx;
if (elements % 2) {
addReply(c,shared.syntaxerr);
return;
}
elements /= 2; /* Now this holds the number of score-element pairs. */
/* Check for incompatible options. */
if (nx && xx) {
addReplyError(c,
"XX and NX options at the same time are not compatible");
return;
}
if (incr && elements > 1) {
addReplyError(c,
"INCR option supports a single increment-element pair");
return;
}
/* Start parsing all the scores, we need to emit any syntax error
* before executing additions to the sorted set, as the command should
* either execute fully or nothing at all. */
scores = zmalloc(sizeof(double)*elements);
for (j = 0; j < elements; j++) {
if (getDoubleFromObjectOrReply(c,c->argv[2+j*2],&scores[j],NULL)
if (getDoubleFromObjectOrReply(c,c->argv[scoreidx+j*2],&scores[j],NULL)
!= REDIS_OK) goto cleanup;
}
/* Lookup the key and create the sorted set if does not exist. */
zobj = lookupKeyWrite(c->db,key);
if (zobj == NULL) {
if (xx) goto reply_to_client; /* No key + XX option: nothing to do. */
if (server.zset_max_ziplist_entries == 0 ||
server.zset_max_ziplist_value < sdslen(c->argv[3]->ptr))
server.zset_max_ziplist_value < sdslen(c->argv[scoreidx+1]->ptr))
{
zobj = createZsetObject();
} else {
@@ -1220,8 +1269,9 @@ void zaddGenericCommand(redisClient *c, int incr) {
unsigned char *eptr;
/* Prefer non-encoded element when dealing with ziplists. */
ele = c->argv[3+j*2];
ele = c->argv[scoreidx+1+j*2];
if ((eptr = zzlFind(zobj->ptr,ele,&curscore)) != NULL) {
if (nx) continue;
if (incr) {
score += curscore;
if (isnan(score)) {
@@ -1237,7 +1287,8 @@ void zaddGenericCommand(redisClient *c, int incr) {
server.dirty++;
updated++;
}
} else {
processed++;
} else if (!xx) {
/* Optimize: check if the element is too large or the list
* becomes too long *before* executing zzlInsert. */
zobj->ptr = zzlInsert(zobj->ptr,ele,score);
@@ -1247,15 +1298,18 @@ void zaddGenericCommand(redisClient *c, int incr) {
zsetConvert(zobj,REDIS_ENCODING_SKIPLIST);
server.dirty++;
added++;
processed++;
}
} else if (zobj->encoding == REDIS_ENCODING_SKIPLIST) {
zset *zs = zobj->ptr;
zskiplistNode *znode;
dictEntry *de;
ele = c->argv[3+j*2] = tryObjectEncoding(c->argv[3+j*2]);
ele = c->argv[scoreidx+1+j*2] =
tryObjectEncoding(c->argv[scoreidx+1+j*2]);
de = dictFind(zs->dict,ele);
if (de != NULL) {
if (nx) continue;
curobj = dictGetKey(de);
curscore = *(double*)dictGetVal(de);
@@ -1280,22 +1334,30 @@ void zaddGenericCommand(redisClient *c, int incr) {
server.dirty++;
updated++;
}
} else {
processed++;
} else if (!xx) {
znode = zslInsert(zs->zsl,score,ele);
incrRefCount(ele); /* Inserted in skiplist. */
redisAssertWithInfo(c,NULL,dictAdd(zs->dict,ele,&znode->score) == DICT_OK);
incrRefCount(ele); /* Added to dictionary. */
server.dirty++;
added++;
processed++;
}
} else {
redisPanic("Unknown sorted set encoding");
}
}
if (incr) /* ZINCRBY */
addReplyDouble(c,score);
else /* ZADD */
addReplyLongLong(c,added);
reply_to_client:
if (incr) { /* ZINCRBY or INCR option. */
if (processed)
addReplyDouble(c,score);
else
addReply(c,shared.nullbulk);
} else { /* ZADD. */
addReplyLongLong(c,ch ? added+updated : added);
}
cleanup:
zfree(scores);
@@ -1307,11 +1369,11 @@ cleanup:
}
void zaddCommand(redisClient *c) {
zaddGenericCommand(c,0);
zaddGenericCommand(c,ZADD_NONE);
}
void zincrbyCommand(redisClient *c) {
zaddGenericCommand(c,1);
zaddGenericCommand(c,ZADD_INCR);
}
void zremCommand(redisClient *c) {
src/version.h
+1 -1
View File
@@ -1 +1 @@
#define REDIS_VERSION "3.0.1"
#define REDIS_VERSION "3.0.4"
tests/integration/rdb.tcl
+13 -13
View File
@@ -7,19 +7,19 @@ start_server [list overrides [list "dir" $server_path "dbfilename" "encodings.rd
test "RDB encoding loading test" {
r select 0
csvdump r
} {"compressible","string","aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa"
"hash","hash","a","1","aa","10","aaa","100","b","2","bb","20","bbb","200","c","3","cc","30","ccc","300","ddd","400","eee","5000000000",
"hash_zipped","hash","a","1","b","2","c","3",
"list","list","1","2","3","a","b","c","100000","6000000000","1","2","3","a","b","c","100000","6000000000","1","2","3","a","b","c","100000","6000000000",
"list_zipped","list","1","2","3","a","b","c","100000","6000000000",
"number","string","10"
"set","set","1","100000","2","3","6000000000","a","b","c",
"set_zipped_1","set","1","2","3","4",
"set_zipped_2","set","100000","200000","300000","400000",
"set_zipped_3","set","1000000000","2000000000","3000000000","4000000000","5000000000","6000000000",
"string","string","Hello World"
"zset","zset","a","1","b","2","c","3","aa","10","bb","20","cc","30","aaa","100","bbb","200","ccc","300","aaaa","1000","cccc","123456789","bbbb","5000000000",
"zset_zipped","zset","a","1","b","2","c","3",
} {"0","compressible","string","aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa"
"0","hash","hash","a","1","aa","10","aaa","100","b","2","bb","20","bbb","200","c","3","cc","30","ccc","300","ddd","400","eee","5000000000",
"0","hash_zipped","hash","a","1","b","2","c","3",
"0","list","list","1","2","3","a","b","c","100000","6000000000","1","2","3","a","b","c","100000","6000000000","1","2","3","a","b","c","100000","6000000000",
"0","list_zipped","list","1","2","3","a","b","c","100000","6000000000",
"0","number","string","10"
"0","set","set","1","100000","2","3","6000000000","a","b","c",
"0","set_zipped_1","set","1","2","3","4",
"0","set_zipped_2","set","100000","200000","300000","400000",
"0","set_zipped_3","set","1000000000","2000000000","3000000000","4000000000","5000000000","6000000000",
"0","string","string","Hello World"
"0","zset","zset","a","1","b","2","c","3","aa","10","bb","20","cc","30","aaa","100","bbb","200","ccc","300","aaaa","1000","cccc","123456789","bbbb","5000000000",
"0","zset_zipped","zset","a","1","b","2","c","3",
}
}
tests/integration/replication-psync.tcl
+38 -25
View File
@@ -13,7 +13,11 @@ proc stop_bg_complex_data {handle} {
#
# You can specifiy backlog size, ttl, delay before reconnection, test duration
# in seconds, and an additional condition to verify at the end.
proc test_psync {descr duration backlog_size backlog_ttl delay cond} {
#
# If reconnect is > 0, the test actually try to break the connection and
# reconnect with the master, otherwise just the initial synchronization is
# checked for consistency.
proc test_psync {descr duration backlog_size backlog_ttl delay cond diskless reconnect} {
start_server {tags {"repl"}} {
start_server {} {
@@ -24,6 +28,8 @@ proc test_psync {descr duration backlog_size backlog_ttl delay cond} {
$master config set repl-backlog-size $backlog_size
$master config set repl-backlog-ttl $backlog_ttl
$master config set repl-diskless-sync $diskless
$master config set repl-diskless-sync-delay 1
set load_handle0 [start_bg_complex_data $master_host $master_port 9 100000]
set load_handle1 [start_bg_complex_data $master_host $master_port 11 100000]
@@ -48,22 +54,24 @@ proc test_psync {descr duration backlog_size backlog_ttl delay cond} {
}
}
test "Test replication partial resync: $descr" {
test "Test replication partial resync: $descr (diskless: $diskless, reconnect: $reconnect)" {
# Now while the clients are writing data, break the maste-slave
# link multiple times.
for {set j 0} {$j < $duration*10} {incr j} {
after 100
# catch {puts "MASTER [$master dbsize] keys, SLAVE [$slave dbsize] keys"}
if ($reconnect) {
for {set j 0} {$j < $duration*10} {incr j} {
after 100
# catch {puts "MASTER [$master dbsize] keys, SLAVE [$slave dbsize] keys"}
if {($j % 20) == 0} {
catch {
if {$delay} {
$slave multi
$slave client kill $master_host:$master_port
$slave debug sleep $delay
$slave exec
} else {
$slave client kill $master_host:$master_port
if {($j % 20) == 0} {
catch {
if {$delay} {
$slave multi
$slave client kill $master_host:$master_port
$slave debug sleep $delay
$slave exec
} else {
$slave client kill $master_host:$master_port
}
}
}
}
@@ -98,18 +106,23 @@ proc test_psync {descr duration backlog_size backlog_ttl delay cond} {
}
}
test_psync {ok psync} 6 1000000 3600 0 {
assert {[s -1 sync_partial_ok] > 0}
}
foreach diskless {no yes} {
test_psync {no reconnection, just sync} 6 1000000 3600 0 {
} $diskless 0
test_psync {no backlog} 6 100 3600 0.5 {
assert {[s -1 sync_partial_err] > 0}
}
test_psync {ok psync} 6 1000000 3600 0 {
assert {[s -1 sync_partial_ok] > 0}
} $diskless 1
test_psync {ok after delay} 3 100000000 3600 3 {
assert {[s -1 sync_partial_ok] > 0}
}
test_psync {no backlog} 6 100 3600 0.5 {
assert {[s -1 sync_partial_err] > 0}
} $diskless 1
test_psync {backlog expired} 3 100000000 1 3 {
assert {[s -1 sync_partial_err] > 0}
test_psync {ok after delay} 3 100000000 3600 3 {
assert {[s -1 sync_partial_ok] > 0}
} $diskless 1
test_psync {backlog expired} 3 100000000 1 3 {
assert {[s -1 sync_partial_err] > 0}
} $diskless 1
}
tests/sentinel/tests/06-ckquorum.tcl
+34
View File
@@ -0,0 +1,34 @@
# Test for the SENTINEL CKQUORUM command
source "../tests/includes/init-tests.tcl"
set num_sentinels [llength $::sentinel_instances]
test "CKQUORUM reports OK and the right amount of Sentinels" {
foreach_sentinel_id id {
assert_match "*OK $num_sentinels usable*" [S $id SENTINEL CKQUORUM mymaster]
}
}
test "CKQUORUM detects quorum cannot be reached" {
set orig_quorum [expr {$num_sentinels/2+1}]
S 0 SENTINEL SET mymaster quorum [expr {$num_sentinels+1}]
catch {[S 0 SENTINEL CKQUORUM mymaster]} err
assert_match "*NOQUORUM*" $err
S 0 SENTINEL SET mymaster quorum $orig_quorum
}
test "CKQUORUM detects failover authorization cannot be reached" {
set orig_quorum [expr {$num_sentinels/2+1}]
S 0 SENTINEL SET mymaster quorum 1
kill_instance sentinel 1
kill_instance sentinel 2
kill_instance sentinel 3
after 5000
catch {[S 0 SENTINEL CKQUORUM mymaster]} err
assert_match "*NOQUORUM*" $err
S 0 SENTINEL SET mymaster quorum $orig_quorum
restart_instance sentinel 1
restart_instance sentinel 2
restart_instance sentinel 3
}
tests/support/util.tcl
+36 -32
View File
@@ -262,46 +262,50 @@ proc formatCommand {args} {
proc csvdump r {
set o {}
foreach k [lsort [{*}$r keys *]] {
set type [{*}$r type $k]
append o [csvstring $k] , [csvstring $type] ,
switch $type {
string {
append o [csvstring [{*}$r get $k]] "\n"
}
list {
foreach e [{*}$r lrange $k 0 -1] {
append o [csvstring $e] ,
for {set db 0} {$db < 16} {incr db} {
{*}$r select $db
foreach k [lsort [{*}$r keys *]] {
set type [{*}$r type $k]
append o [csvstring $db] , [csvstring $k] , [csvstring $type] ,
switch $type {
string {
append o [csvstring [{*}$r get $k]] "\n"
}
append o "\n"
}
set {
foreach e [lsort [{*}$r smembers $k]] {
append o [csvstring $e] ,
list {
foreach e [{*}$r lrange $k 0 -1] {
append o [csvstring $e] ,
}
append o "\n"
}
append o "\n"
}
zset {
foreach e [{*}$r zrange $k 0 -1 withscores] {
append o [csvstring $e] ,
set {
foreach e [lsort [{*}$r smembers $k]] {
append o [csvstring $e] ,
}
append o "\n"
}
append o "\n"
}
hash {
set fields [{*}$r hgetall $k]
set newfields {}
foreach {k v} $fields {
lappend newfields [list $k $v]
zset {
foreach e [{*}$r zrange $k 0 -1 withscores] {
append o [csvstring $e] ,
}
append o "\n"
}
set fields [lsort -index 0 $newfields]
foreach kv $fields {
append o [csvstring [lindex $kv 0]] ,
append o [csvstring [lindex $kv 1]] ,
hash {
set fields [{*}$r hgetall $k]
set newfields {}
foreach {k v} $fields {
lappend newfields [list $k $v]
}
set fields [lsort -index 0 $newfields]
foreach kv $fields {
append o [csvstring [lindex $kv 0]] ,
append o [csvstring [lindex $kv 1]] ,
}
append o "\n"
}
append o "\n"
}
}
}
{*}$r select 9
return $o
}
tests/unit/type/set.tcl
+1
View File
@@ -450,6 +450,7 @@ start_server {
test "SMOVE non existing key" {
setup_move
assert_equal 0 [r smove myset1 myset2 foo]
assert_equal 0 [r smove myset1 myset1 foo]
assert_equal {1 a b} [lsort [r smembers myset1]]
assert_equal {2 3 4} [lsort [r smembers myset2]]
}
tests/unit/type/zset.tcl
+80
View File
@@ -43,6 +43,84 @@ start_server {tags {"zset"}} {
assert_error "*not*float*" {r zadd myzset nan abc}
}
test "ZADD with options syntax error with incomplete pair" {
r del ztmp
catch {r zadd ztmp xx 10 x 20} err
set err
} {ERR*}
test "ZADD XX option without key - $encoding" {
r del ztmp
assert {[r zadd ztmp xx 10 x] == 0}
assert {[r type ztmp] eq {none}}
}
test "ZADD XX existing key - $encoding" {
r del ztmp
r zadd ztmp 10 x
assert {[r zadd ztmp xx 20 y] == 0}
assert {[r zcard ztmp] == 1}
}
test "ZADD XX returns the number of elements actually added" {
r del ztmp
r zadd ztmp 10 x
set retval [r zadd ztmp 10 x 20 y 30 z]
assert {$retval == 2}
}
test "ZADD XX updates existing elements score" {
r del ztmp
r zadd ztmp 10 x 20 y 30 z
r zadd ztmp xx 5 foo 11 x 21 y 40 zap
assert {[r zcard ztmp] == 3}
assert {[r zscore ztmp x] == 11}
assert {[r zscore ztmp y] == 21}
}
test "ZADD XX and NX are not compatible" {
r del ztmp
catch {r zadd ztmp xx nx 10 x} err
set err
} {ERR*}
test "ZADD NX with non exisitng key" {
r del ztmp
r zadd ztmp nx 10 x 20 y 30 z
assert {[r zcard ztmp] == 3}
}
test "ZADD NX only add new elements without updating old ones" {
r del ztmp
r zadd ztmp 10 x 20 y 30 z
assert {[r zadd ztmp nx 11 x 21 y 100 a 200 b] == 2}
assert {[r zscore ztmp x] == 10}
assert {[r zscore ztmp y] == 20}
assert {[r zscore ztmp a] == 100}
assert {[r zscore ztmp b] == 200}
}
test "ZADD INCR works like ZINCRBY" {
r del ztmp
r zadd ztmp 10 x 20 y 30 z
r zadd ztmp INCR 15 x
assert {[r zscore ztmp x] == 25}
}
test "ZADD INCR works with a single score-elemenet pair" {
r del ztmp
r zadd ztmp 10 x 20 y 30 z
catch {r zadd ztmp INCR 15 x 10 y} err
set err
} {ERR*}
test "ZADD CH option changes return value to all changed elements" {
r del ztmp
r zadd ztmp 10 x 20 y 30 z
assert {[r zadd ztmp 11 x 21 y 30 z] == 0}
assert {[r zadd ztmp ch 12 x 22 y 30 z] == 2}
}
test "ZINCRBY calls leading to NaN result in error" {
r zincrby myzset +inf abc
assert_error "*NaN*" {r zincrby myzset -inf abc}
@@ -77,6 +155,8 @@ start_server {tags {"zset"}} {
}
test "ZCARD basics - $encoding" {
r del ztmp
r zadd ztmp 10 a 20 b 30 c
assert_equal 3 [r zcard ztmp]
assert_equal 0 [r zcard zdoesntexist]
}