01:41:36 GMT <wizwizwizw> so can someone help me understand (or point me to a good article) on why i shouldn't just use a hashmap in memory rather than redis?
01:41:54 GMT <wizwizwizw> particularly if i just have a single process application,...
05:32:42 GMT <kaos01_> i got sentinel setup they recognise master being down but no failover happens
05:32:59 GMT <kaos01_> all 3 sentinals show +sdown master mymaster
05:33:20 GMT <kaos01_> -sdown master mymaster
06:18:29 GMT <cloudbud> how can i change the max memory configuration at run time
06:18:38 GMT <cloudbud> withour restarting redis server
06:33:12 GMT <minus> cloudbud: config set maxmemory (works for most config options)
06:33:26 GMT <minus> config get * to read the config
06:33:53 GMT <cloudbud> minus : i want to set the max memory to 16 GB right now it is 20000000000 bytes what would be the command then
06:34:18 GMT <cloudbud> minus : config get * gives
06:34:22 GMT <cloudbud>  13) "maxmemory"
06:34:26 GMT <cloudbud>   14) "20000000000"
06:35:23 GMT <minus> just config set maxmemory 16000000000
06:35:30 GMT <cloudbud> my redis version is redis_version:2.8.4
06:35:33 GMT <cloudbud> so will it work
06:36:27 GMT <minus> probably, i don't know
06:53:48 GMT <kaos01_> my slave has master0:name=mymaster,status=sdown while otehr two sentinels odown
08:30:11 GMT <htd> hi guys, if I disable persistance in the config (i.e. no save" setting) but call BGSAVE say once a day from my client, it will still write the RDB file, correct?
08:30:16 GMT <htd> (trying to see if my plans for a embedded db that should not write much to SD card could work)
08:33:02 GMT <methodism> how can i check redis version?
08:41:12 GMT <minus> methodism: the INFO command
08:41:18 GMT <methodism> thx
08:57:03 GMT <du5tball> if i run a re
08:57:39 GMT <du5tball> if i run a redis-cluster and want it to more or less auto-manage, do i set up the sentinel and masters, but not run the redis-trib.rb script?
09:04:36 GMT <badboy_> don't mix Redis Cluster and Sentinel
09:13:58 GMT <King_DuckZ> hi, this is probably due to me being a lua noob, but I can't get this to work https://alarmpi.no-ip.org/kamokan/bd?lua when the desired token is not in redis, hmget returns nil, but I can't get the condition at line 3 to ever be met, I tried not result, nil != result[1], result.err...
09:21:06 GMT <badboy_> if result == false
09:23:35 GMT <King_DuckZ> -_-
09:23:40 GMT <King_DuckZ> badboy_: thanks
09:24:09 GMT <badboy_> > Redis Nil bulk reply and Nil multi bulk reply -> Lua false boolean type
09:24:14 GMT <badboy_> from https://redis.io/commands/eval#conversion-between-lua-and-redis-data-types
09:25:39 GMT <du5tball> can someone explain to me what the sentinel actually does? i'm reading confusing information. initially i thought the sentinel is basically an auto-cluster, but that doesn't seem to be the case as i still need the actual redis-server running. then i assumed the sentinel does management for an existing redis-cluster, but i've read that you shouldn't point sentinel at a cluster
09:25:40 GMT <du5tball> (https://fnordig.de/2015/06/01/redis-sentinel-and-redis-cluster/)
09:26:07 GMT <badboy_> Sentinel = Monitor & Failover Redis nodes
09:26:28 GMT <badboy_> Cluster = full-blown sharded cluster solution with automatic failover
09:27:06 GMT <du5tball> badboy_: does the sentinel replicate between non-clustered nodes?
09:27:10 GMT <King_DuckZ> badboy_: I know that page, I was focusing on the error reply and the table with err... btw it doesn't look like it's working
09:27:29 GMT <badboy_> du5tball: Sentinel _only_ monitors existing Redis master nodes and their slaves
09:27:32 GMT <King_DuckZ> I still get a nil/0/0/nil reply if I eval that :(
09:27:45 GMT <badboy_> du5tball: Sentinel has nothing to do with the actual data stored
09:27:49 GMT <du5tball> okay
09:27:56 GMT <du5tball> back in a bit, getting dragged to lunch
09:28:58 GMT <badboy_> King_DuckZ: use the debugger
09:32:01 GMT <King_DuckZ> badboy_: I didn't know there was one! should I just look for "lua debugger" or do you have a name for a good one? btw I'll look into that later because I just got to the office and I can't really debug my stuff right now :/
09:32:12 GMT <badboy_> https://redis.io/topics/ldb
09:33:41 GMT <King_DuckZ> thanks
10:20:14 GMT <du5tball> badboy_: is there another way besides the ruby-script to setup the cluster?
10:21:59 GMT <badboy_> manually
10:22:20 GMT <badboy_> there was at least one other tool around. basically a reimplementation of redis-trib, but I forgot itsname
10:23:39 GMT <du5tball> is there a guide to manually?
10:24:16 GMT <badboy_> not really. there's a tutorial: https://redis.io/topics/cluster-tutorial - but it relies on redis-trib
10:24:24 GMT <du5tball> that's what i found as well
10:24:27 GMT <badboy_> the best you can do is read the https://redis.io/topics/cluster-spec or the code
10:24:40 GMT <badboy_> It's not that hard.
10:24:45 GMT <badboy_> Why do you want to avoid redis-trib?
10:25:00 GMT <du5tball> badboy_: because i'm using sles
10:25:08 GMT <badboy_> (there are plans to rewrite redis-trib in C and include it in the general redis package)
10:25:32 GMT <du5tball> hm, the server has ruby 2.1 available
10:25:44 GMT <badboy_> everything >= 1.9 should do.
10:25:58 GMT <du5tball> okay. i'll try
11:09:38 GMT <_Wise__> hi *
11:10:35 GMT <_Wise__> is it possible to trigger a Redis 3.2.8 failover when not in Cluster mode ?
11:15:14 GMT <_Wise__> I'm looking for a way to gracefully failover master -> slave
11:15:44 GMT <_Wise__> I have 1 master, 1 slave and 3 sentinels
11:21:21 GMT <du5tball> what happens when i "cluster" by sharing the db and AOF via NFS between several standalone redis instances?
11:34:09 GMT <_Wise__> hmm, "sentinel failover <master-group-name>" seems to do the trick
11:37:09 GMT <du5tball> after thinking about it, i retract my question
17:54:24 GMT <devslash> is redis considered a distributed or local cache
18:35:37 GMT <devon> have a question.. this issue just came up.. I’m looking at the cluster status and I see one slace node in a fail state.. I restarted the service and logs are good (they were good before the restart). I’m not sure why its in a failed state
20:33:38 GMT <King_DuckZ> hi, I'm a bit confused by binary data in redis, when I write data from incredis as RedisInt, 10 in my example, from redis-cli I see it as "\x00\x00\x00\x00\x00\x00\x00\n"
20:34:07 GMT <King_DuckZ> that is 10 in binary, but it's not really what I expected :/
20:35:30 GMT <King_DuckZ> in fact when I try to set the expiry as that I get an error, should I always pass strings around instead?
20:44:57 GMT <King_DuckZ> ok I think I fixed it, I'm just misunderstanding stuff
20:56:56 GMT <minus> use --raw with redis-cli, then redis-cli won't escape it
20:57:25 GMT <minus> the \n is weird tho, you must be writing that superfluously
20:59:53 GMT <King_DuckZ> minus: that's 10
20:59:59 GMT <King_DuckZ> in ascii at least
21:03:10 GMT <minus> ohhh
21:03:16 GMT <minus> my mistake
21:03:53 GMT <minus> redis-cli --raw GET the-key | hexdump -C should show you that it's fine
22:57:55 GMT <kaos01> im testing failover, if i stop the master, replica becomes mastar - all good, but than i start up master, stop replicat(the new master) and all fails
22:58:22 GMT <kaos01> end up with replica down (as i stoped it) master readonly (as failback fails)