]>
git.saurik.com Git - redis.git/log
Damian Janowski & Michel Martens [Tue, 9 Nov 2010 13:31:02 +0000 (10:31 -0300)]
Handle BRPOPLPUSH inside a transaction.
Damian Janowski & Michel Martens [Mon, 8 Nov 2010 23:47:46 +0000 (20:47 -0300)]
Refactor code for BRPOPLPUSH.
Damian Janowski & Michel Martens [Mon, 8 Nov 2010 22:38:01 +0000 (19:38 -0300)]
Move to struct.
Damian Janowski & Michel Martens [Mon, 8 Nov 2010 18:25:59 +0000 (15:25 -0300)]
BRPOPLPUSH.
Damian Janowski & Michel Martens [Mon, 8 Nov 2010 13:43:21 +0000 (10:43 -0300)]
Fix case in RPOPLPUSH.
antirez [Mon, 29 Nov 2010 11:20:17 +0000 (12:20 +0100)]
quick and dirty fix for hiredis bug creating problem with the new redis-cli connect commmand. Also change prompt when redis-cli is not connected
antirez [Mon, 29 Nov 2010 11:17:55 +0000 (12:17 +0100)]
new redis-cli command connect
antirez [Mon, 29 Nov 2010 10:14:57 +0000 (11:14 +0100)]
gitignore now includes a few more files
Pieter Noordhuis [Sun, 21 Nov 2010 15:44:17 +0000 (16:44 +0100)]
Don't hardcode make to "make"
Pieter Noordhuis [Sun, 21 Nov 2010 15:27:47 +0000 (16:27 +0100)]
Stop using /dev/null to find out the serialized object length
Now the rdbSave* functions return the number of bytes written (or
required to write) in serializing a Redis object, writing to /dev/null
and using ftell (which doesn't work on FreeBSD) isn't needed anymore.
Pieter Noordhuis [Sun, 21 Nov 2010 15:12:25 +0000 (16:12 +0100)]
Wrapper around fwrite to make API consistent
Pieter Noordhuis [Sun, 21 Nov 2010 14:39:34 +0000 (15:39 +0100)]
Return number of bytes written from rdbSave* functions
antirez [Mon, 15 Nov 2010 14:50:41 +0000 (15:50 +0100)]
added a few more files to gitignore
antirez [Fri, 12 Nov 2010 19:02:20 +0000 (20:02 +0100)]
Fix for bug 374, thanks to Jeremy Zawodny for reporting and tracing why it was crashing.
antirez [Thu, 11 Nov 2010 12:19:17 +0000 (13:19 +0100)]
volatile-lru maxmemory policy segfault fixed, thanks to Anthony Lauzon for reporting the problem with the patch. Original patch modified a bit in order to avoid the double lookup if the policy is allkeys-lru
antirez [Mon, 8 Nov 2010 15:26:02 +0000 (16:26 +0100)]
now redis-cli is able to show the Git SHA1 in the version output
antirez [Mon, 8 Nov 2010 15:14:15 +0000 (16:14 +0100)]
make sure to flush stdout every line read in monitor mode, to play well with redirection to file
antirez [Mon, 8 Nov 2010 15:12:16 +0000 (16:12 +0100)]
added noeviction policy to redis maxmemory. ZSCORE removed from the list of commands that can't be called when we are low on memory, this command was added in the past for a stupid error.
antirez [Mon, 8 Nov 2010 12:19:58 +0000 (13:19 +0100)]
build redis-server at the end so have a more pleasing to see Make output and the advice to run the test suite at the end.
antirez [Mon, 8 Nov 2010 11:53:36 +0000 (12:53 +0100)]
fixed compilation with 32bit target
antirez [Mon, 8 Nov 2010 10:52:03 +0000 (11:52 +0100)]
non blocking loading of the DB / AOF with informations and ETA in INFO output
Pieter Noordhuis [Fri, 5 Nov 2010 16:24:48 +0000 (17:24 +0100)]
Update hiredis
Pieter Noordhuis [Fri, 5 Nov 2010 16:22:16 +0000 (17:22 +0100)]
Only cascade clean target to deps
antirez [Fri, 5 Nov 2010 10:00:20 +0000 (11:00 +0100)]
version bumped to 2.1.7 after merging with aaslave branch for non blocking slaves
antirez [Fri, 5 Nov 2010 09:59:49 +0000 (10:59 +0100)]
Merge branch 'aaslave'
antirez [Fri, 5 Nov 2010 09:57:53 +0000 (10:57 +0100)]
version bumped to 2.1.6
antirez [Fri, 5 Nov 2010 09:29:59 +0000 (10:29 +0100)]
Merge remote branch 'pietern/benchmark'
Pieter Noordhuis [Fri, 5 Nov 2010 09:22:52 +0000 (10:22 +0100)]
Use multi-bulk protocol in SET benchmark
antirez [Thu, 4 Nov 2010 18:59:21 +0000 (19:59 +0100)]
config option to select if when replication link with master a slave should or not serve stale data
antirez [Thu, 4 Nov 2010 17:50:23 +0000 (18:50 +0100)]
replication asynchronous SYNC information in INFO output
antirez [Thu, 4 Nov 2010 17:14:20 +0000 (18:14 +0100)]
more replication info in logs
antirez [Thu, 4 Nov 2010 17:09:35 +0000 (18:09 +0100)]
non blocking slave replication is now more non blocking than the first implementation...
antirez [Thu, 4 Nov 2010 16:35:03 +0000 (17:35 +0100)]
typos and minor stuff fixed in the new non blocking replication code
antirez [Thu, 4 Nov 2010 16:29:53 +0000 (17:29 +0100)]
first attempt to non blocking implementation of slave replication and SYNC bulk data download. Never compiled so far...
Pieter Noordhuis [Thu, 4 Nov 2010 13:47:15 +0000 (14:47 +0100)]
Microsecond latency resolution in redis-benchmark
Pieter Noordhuis [Thu, 4 Nov 2010 12:37:05 +0000 (13:37 +0100)]
Use hiredis from redis-benchmark
Pieter Noordhuis [Thu, 4 Nov 2010 12:35:47 +0000 (13:35 +0100)]
Update hiredis
antirez [Thu, 4 Nov 2010 09:48:49 +0000 (10:48 +0100)]
minor test suite bug fixed
Pieter Noordhuis [Thu, 4 Nov 2010 09:09:30 +0000 (10:09 +0100)]
Stop using the freelist robj* cache
antirez [Wed, 3 Nov 2010 17:14:09 +0000 (18:14 +0100)]
top level make clean also clean hiredis and linoise (deps)
Pieter Noordhuis [Wed, 3 Nov 2010 16:10:12 +0000 (17:10 +0100)]
Import linenoise as dependency for redis-cli
Pieter Noordhuis [Wed, 3 Nov 2010 16:07:10 +0000 (17:07 +0100)]
Put duration in parenthesis
Pieter Noordhuis [Wed, 3 Nov 2010 16:03:54 +0000 (17:03 +0100)]
Add proper numbering for multi bulk replies in redis-cli
Pieter Noordhuis [Wed, 3 Nov 2010 15:09:38 +0000 (16:09 +0100)]
Use hiredis from redis-cli
Pieter Noordhuis [Wed, 3 Nov 2010 15:03:04 +0000 (16:03 +0100)]
Add hiredis dependency for redis-cli, redis-benchmark, etc
antirez [Wed, 3 Nov 2010 11:14:36 +0000 (12:14 +0100)]
added support for command renaming/suppression in redis.conf
antirez [Wed, 3 Nov 2010 10:23:59 +0000 (11:23 +0100)]
command lookup process turned into a much more flexible and probably faster hash table
antirez [Wed, 3 Nov 2010 09:31:19 +0000 (10:31 +0100)]
Now it is possible to use quoted strings in the redis.conf file accordingly to the sdssplitargs() function capabilities.
antirez [Tue, 2 Nov 2010 22:47:52 +0000 (23:47 +0100)]
Merge remote branch 'pietern/unixsocket'
antirez [Tue, 2 Nov 2010 21:47:10 +0000 (22:47 +0100)]
RSS information in INFO output
antirez [Tue, 2 Nov 2010 17:59:48 +0000 (18:59 +0100)]
Do not update the LRU info on key lookup when we have a saving child. With this trivial change the additional memory used while saving with a background child in presence of many read operations is zero.
antirez [Tue, 2 Nov 2010 17:08:30 +0000 (18:08 +0100)]
redis-cli in interactive mode now prints the time elapsed of the operation performed took more than half a second.
antirez [Tue, 2 Nov 2010 11:10:41 +0000 (12:10 +0100)]
Revert "Function to compute RSS memory usage in a fast way, suitable to be called inside keys eviction loops"
This reverts commit
7d47ecd543ed41db70396bc3866eac872a31774d .
antirez [Tue, 2 Nov 2010 11:09:59 +0000 (12:09 +0100)]
Revert "Now maxmemory, VM, and everything else uses the fast RSS memory used estimation instead of raw memory reported by zmalloc(). This means that setting maxmemory to 2GB will really have the effect of using up to 2GB of memory."
This reverts commit
a3e60027e7d4419971406c6da7d6e8accbcaf9b0 .
antirez [Tue, 2 Nov 2010 10:50:55 +0000 (11:50 +0100)]
Now maxmemory, VM, and everything else uses the fast RSS memory used estimation instead of raw memory reported by zmalloc(). This means that setting maxmemory to 2GB will really have the effect of using up to 2GB of memory.
antirez [Tue, 2 Nov 2010 10:40:35 +0000 (11:40 +0100)]
Function to compute RSS memory usage in a fast way, suitable to be called inside keys eviction loops
antirez [Tue, 2 Nov 2010 10:15:09 +0000 (11:15 +0100)]
removed a number of stupid compilation warnings on Linux
antirez [Tue, 2 Nov 2010 09:51:09 +0000 (10:51 +0100)]
zmalloc functions to get RSS and fragmentation refactored into two separated functions
antirez [Thu, 28 Oct 2010 20:59:47 +0000 (22:59 +0200)]
merge conflict resolved
Pieter Noordhuis [Thu, 28 Oct 2010 15:59:05 +0000 (16:59 +0100)]
Rewrite comment that was no longer valid
Pieter Noordhuis [Thu, 28 Oct 2010 15:52:16 +0000 (16:52 +0100)]
When REDIS_CLOSE_AFTER_REPLY is set, there may never be new replies
Pieter Noordhuis [Thu, 28 Oct 2010 14:07:45 +0000 (15:07 +0100)]
Unify two client flags that mean the same
antirez [Thu, 28 Oct 2010 12:12:25 +0000 (14:12 +0200)]
Merge remote branch 'remotes/pietern/zrevrangebyscore'
antirez [Wed, 27 Oct 2010 15:11:17 +0000 (17:11 +0200)]
removed useless spaces from DEBUG OBJECT output
Pieter Noordhuis [Tue, 26 Oct 2010 10:26:27 +0000 (12:26 +0200)]
Return error to client on wrong type for HMGET
antirez [Mon, 25 Oct 2010 08:54:37 +0000 (10:54 +0200)]
added a missing prototype from syncio.c in redis.h
antirez [Mon, 25 Oct 2010 08:49:15 +0000 (10:49 +0200)]
more generally usable i/o functions moved to syncio.c
antirez [Sun, 24 Oct 2010 14:22:52 +0000 (16:22 +0200)]
synchronous I/O networking functions originally used just for replication refactored in a file as generally useful, they are used in the cluster branch for MIGRATE.
Pieter Noordhuis [Sat, 23 Oct 2010 08:18:48 +0000 (10:18 +0200)]
Exclusively use either tcmalloc or OSX's native malloc_size()
Pieter Noordhuis [Sat, 23 Oct 2010 07:59:28 +0000 (09:59 +0200)]
Don't use prefix when malloc_size() can be called
Also, use tcmalloc functions explicitly via macros to prevent symbol
lookups to resolve to native malloc/free on OSX.
antirez [Fri, 22 Oct 2010 21:30:48 +0000 (23:30 +0200)]
Make sure to reset the signal handler and deliver again the original crashing signal when dumping the stack trace. This will allow to dump the core if core dumping is enabled.
Robey Pointer [Thu, 30 Sep 2010 18:53:34 +0000 (11:53 -0700)]
if server.saveparamslen is not set, don't save the DB on exit.
antirez [Thu, 21 Oct 2010 22:18:26 +0000 (00:18 +0200)]
README more specific about tcmalloc space benefits
antirez [Thu, 21 Oct 2010 22:16:32 +0000 (00:16 +0200)]
Makefile typo fixed for tcmalloc option
antirez [Thu, 21 Oct 2010 22:10:17 +0000 (00:10 +0200)]
reports if tcmalloc is in used in INFO output
antirez [Thu, 21 Oct 2010 22:06:44 +0000 (00:06 +0200)]
support for compiling with tcmalloc
antirez [Thu, 21 Oct 2010 15:55:44 +0000 (17:55 +0200)]
version bumped to 2.1.5
Pieter Noordhuis [Sun, 17 Oct 2010 16:09:23 +0000 (18:09 +0200)]
Skip object encoding where it doesn't make sense
Pieter Noordhuis [Sun, 17 Oct 2010 15:46:55 +0000 (17:46 +0200)]
Object encoding in hash function is done by a more specific function
Pieter Noordhuis [Sun, 17 Oct 2010 15:31:40 +0000 (17:31 +0200)]
Removed unused command flags
Pieter Noordhuis [Sun, 17 Oct 2010 15:21:41 +0000 (17:21 +0200)]
Convert objects in the command procs instead of the protocol code
Pieter Noordhuis [Fri, 15 Oct 2010 17:15:38 +0000 (19:15 +0200)]
Prevent clients from making too large multibulk requests
Pieter Noordhuis [Fri, 15 Oct 2010 16:17:06 +0000 (18:17 +0200)]
Add benchmark for MSET
antirez [Fri, 15 Oct 2010 16:04:05 +0000 (18:04 +0200)]
prevent small integer sharing when maxmemory is active. So every object will use a private LRU field and the LRU algorithm can work well
Pieter Noordhuis [Fri, 15 Oct 2010 15:34:20 +0000 (17:34 +0200)]
Change protocol from bulk to inline in redis-benchmark
Pieter Noordhuis [Fri, 15 Oct 2010 15:27:05 +0000 (17:27 +0200)]
Don't reset the client when processCommand returns REDIS_ERR
Pieter Noordhuis [Fri, 15 Oct 2010 15:25:20 +0000 (17:25 +0200)]
Change tests to use either the inline or the multibulk protocol
Pieter Noordhuis [Fri, 15 Oct 2010 13:56:16 +0000 (15:56 +0200)]
Show output of leaks command on a leak
Pieter Noordhuis [Fri, 15 Oct 2010 13:50:29 +0000 (15:50 +0200)]
Change tcl client to only use the multibulk protocol
Pieter Noordhuis [Fri, 15 Oct 2010 13:40:25 +0000 (15:40 +0200)]
Refactor request parsing code for efficiency
Pieter Noordhuis [Fri, 15 Oct 2010 10:54:53 +0000 (12:54 +0200)]
Add tests for OK on QUIT
antirez [Fri, 15 Oct 2010 10:29:05 +0000 (12:29 +0200)]
CONFIG RESETSTAT no longer resets the server uptime. Now keyspace hits/misses are reset as well.
antirez [Fri, 15 Oct 2010 10:22:48 +0000 (12:22 +0200)]
minor aesthetic change
antirez [Fri, 15 Oct 2010 10:19:21 +0000 (12:19 +0200)]
added keyspace_hits and keyspace_misses fields in INFO output
antirez [Fri, 15 Oct 2010 09:57:38 +0000 (11:57 +0200)]
maxmemory-samples implemented in CONFIG command and configuration file
antirez [Fri, 15 Oct 2010 09:29:03 +0000 (11:29 +0200)]
Don't increment dirty on expireIfNeeded() as natural expires are not considered database changes. This will avoid useless read-only commands in the AOF file as a result of, for instance, a GET operation triggering an expirIfNeeded() call resulting in an expired key removed.
antirez [Thu, 14 Oct 2010 19:23:01 +0000 (21:23 +0200)]
Merge remote branch 'pietern/ziplist-regression'
antirez [Thu, 14 Oct 2010 19:22:21 +0000 (21:22 +0200)]
implemented different algorithms for maxmemory
Pieter Noordhuis [Thu, 14 Oct 2010 19:11:42 +0000 (21:11 +0200)]
Replace ziplist stresser and fix regression
antirez [Thu, 14 Oct 2010 11:52:58 +0000 (13:52 +0200)]
Object approximated LRU algorithm enhanced / fixed / refactored. This is used for the VM currently but will soon be used for maxmemory expiring.