- 04 May, 2015 2 commits
-
-
antirez authored
Should not be an issue given that the precision is the second here, at least if we are using a decent HZ value and the cached time refreshes enough times. So the cached time is only used if HZ is >= 10.
-
antirez authored
Commands may use cached time when the precision is not vital. It is a good idea to refresh it from time to time during the execution of long running scripts or transactions composed of quite a lot of commands.
-
- 30 Apr, 2015 1 commit
-
-
antirez authored
mstime() is not going to change significantly every 16 iterations, so now we update `now` with the current milliseconds time only when we check if the time limit was reached.
-
- 29 Apr, 2015 5 commits
-
-
antirez authored
When the key expires far in the future compared to the cached time in server.mstime, calling mstime(), that calls gettimeofday(), should not be very useful. Instead when we are near the expire, we want the additional precision. This commit is related to issue #2552.
-
antirez authored
This commit allows to avoid two mstime() calls inside the call() function, when the following conditions are true: 1. slowlog is disabled. 2. latency monitoring is disabled. 3. command time acconuting is disabled. Note that '3' was not configurable, this patch just disable it without really allowing the user to turn it on, since this is currently an experiment. If the commit will be merged into unstable, proper support to configure this parameter will be added. Related to issue #2552.
-
antirez authored
-
antirez authored
As suggested in #2543.
-
Salvatore Sanfilippo authored
Fix spelling and grammatical errors in readme
-
- 28 Apr, 2015 3 commits
-
-
antirez authored
This fixes issue #2535, that was actually an hiredis library bug (I submitted an issue and fix to the redis/hiredis repo as well). When an asynchronous hiredis connection subscribes to a Pub/Sub channel and gets an error, and in other related conditions, the function redisProcessCallbacks() enters a code path where the link is disconnected, however the function returns before freeing the allocated reply object. This causes a memory leak. The memory leak was trivial to trigger in Redis Sentinel, which uses hiredis, every time we tried to subscribe to an instance that required a password, in case the Sentinel was configured either with the wrong password or without password at all. In this case, the -AUTH error caused the leaking code path to be executed. It was verified with Valgrind that after this change the leak no longer happens in Sentinel with a misconfigured authentication password.
-
Jan-Erik Rediger authored
-
Jan-Erik Rediger authored
Closes #2549
-
- 27 Apr, 2015 2 commits
- 26 Apr, 2015 1 commit
-
-
antirez authored
An user changed the behavior via a PR without upgrading the doc.
-
- 24 Apr, 2015 1 commit
-
-
Salvatore Sanfilippo authored
Added reference to IANA ticket for port 6379
-
- 23 Apr, 2015 1 commit
-
-
Itamar Haber authored
Just so it's extra official
😄
-
- 01 Apr, 2015 2 commits
-
-
antirez authored
When we fail to setup the write handler it does not make sense to take the client around, it is missing writes: whatever is a client or a slave anyway the connection should terminated ASAP. Moreover what the function does exactly with its return value, and in which case the write handler is installed on the socket, was not clear, so the functions comment are improved to make the goals of the function more obvious. Also related to #2485.
-
Salvatore Sanfilippo authored
fixes to diskless replication.
-
- 31 Mar, 2015 3 commits
-
-
antirez authored
This should likely fix a false positive when running with the --valgrind option.
-
Oran Agra authored
master was closing the connection if the RDB transfer took long time. and also sent PINGs to the slave before it got the initial ACK, in which case the slave wouldn't be able to find the EOF marker.
-
antirez authored
Segfault introduced during a refactoring / warning suppression a few commits away. This particular call assumed that it is safe to pass NULL to the object pointer argument when we are sure the set has a given encoding. This can't be assumed and is now guaranteed to segfault because of the new API of setTypeNext().
-
- 30 Mar, 2015 4 commits
-
-
antirez authored
-
antirez authored
This change fixes several warnings compiling at -O3 level with GCC 4.8.2, and at the same time, in case of misuse of the API, we have the pointer initialize to NULL or the integer initialized to the value -123456789 which is easy to spot by naked eye.
-
antirez authored
Another one just to avoid a warning. Slightly more defensive code anyway.
-
antirez authored
Change done in order to remove a warning and improve code robustness. No actual bug here.
-
- 27 Mar, 2015 3 commits
-
-
antirez authored
-
antirez authored
No semantical changes since to make dict.c truly able to scale over the 32 bit table size limit, the hash function shoulds and other internals related to hash function output should be 64 bit ready.
-
antirez authored
rehashidx is always positive in the two code paths, since the only negative value it could have is -1 when there is no rehashing in progress, and the condition is explicitly checked.
-
- 24 Mar, 2015 3 commits
-
-
antirez authored
Bug as old as Redis and blocking operations. It's hard to trigger since only happens on instance role switch, but the results are quite bad since an inconsistency between master and slave is created. How to trigger the bug is a good description of the bug itself. 1. Client does "BLPOP mylist 0" in master. 2. Master is turned into slave, that replicates from New-Master. 3. Client does "LPUSH mylist foo" in New-Master. 4. New-Master propagates write to slave. 5. Slave receives the LPUSH, the blocked client get served. Now Master "mylist" key has "foo", Slave "mylist" key is empty. Highlights: * At step "2" above, the client remains attached, basically escaping any check performed during command dispatch: read only slave, in that case. * At step "5" the slave (that was the master), serves the blocked client consuming a list element, which is not consumed on the master side. This scenario is technically likely to happen during failovers, however since Redis Sentinel already disconnects clients using the CLIENT command when changing the role of the instance, the bug is avoided in Sentinel deployments. Closes #2473.
-
antirez authored
There was a bug in Redis Cluster caused by clients blocked in a blocking list pop operation, for keys no longer handled by the instance, or in a condition where the cluster became down after the client blocked. A typical situation is: 1) BLPOP <somekey> 0 2) <somekey> hash slot is resharded to another master. The client will block forever int this case. A symmentrical non-cluster-specific bug happens when an instance is turned from master to slave. In that case it is more serious since this will desynchronize data between slaves and masters. This other bug was discovered as a side effect of thinking about the bug explained and fixed in this commit, but will be fixed in a separated commit.
-
Salvatore Sanfilippo authored
create-cluster fix for stop and watch commands
-
- 23 Mar, 2015 1 commit
-
-
superlogical authored
-
- 22 Mar, 2015 5 commits
- 21 Mar, 2015 3 commits