00:23:24 GMT <rfer> oh man what a way to find missed redis server upgrades
00:23:44 GMT <rfer> lua script complaining about incorrect number of args to redis call, which usually means something i passed in was null or an empty table
00:24:15 GMT <rfer> turns out i'm on <2.6.12 on a minority of machines and the format of SET that i'm using isn't available until 2.6.12
01:19:33 GMT <rawburt> Can someone help me with the Python client? I want to search for:   ip = 1.2.3.4 ,  time > 1487207960
02:05:40 GMT <rawburt> Anyone have any experience with Redis-py?
03:17:20 GMT <rfer> rawburt: That's an incredibly vague question. What and how exactly are you searching?
03:44:25 GMT <rawburt> I was looking for a way to search on keys in my value
03:44:51 GMT <rawburt> key: {"foo":"bar","abc":"cde"} for example
06:07:44 GMT <minus> rawburt: redis doesn't really support searching
06:08:18 GMT <minus> rfer: that's one ancient redis version, you should upgrade
06:08:53 GMT <rawburt> minus: Gotcha thanks
06:18:13 GMT <rfer> yeah we are minus, was just disappointed to find out that we don't have matching redis versions across our infrastructure
06:19:17 GMT <minus> Ouch
12:02:10 GMT <_Wise_> hi *
12:02:42 GMT <_Wise_> is there a way in Redis 3.0 to avoid EVALSHA scripts to be logged in the appendonly file ?
12:29:17 GMT <badboy_> _Wise_: No, upgrade to 3.2
12:42:16 GMT <_Wise_> thx badboy_
13:46:55 GMT <michal88> Hi, I have problem with redis 3.2 and sentinel failover on ubuntu14.04 anyone working with that setup?
13:53:00 GMT <_Wise_> michal88: I do this with 3.0, if that can help ?
13:59:38 GMT <michal88> I can't see the failover happening in my setup, I have the same configuration running in ubuntu 16.04 and it's fine, but with the 14.04 I don't get the failover on sentinel
14:00:03 GMT <michal88> it detects the master going down but doesn't run the election
14:00:55 GMT <michal88> This is my sentinel config:
14:00:58 GMT <michal88> sentinel monitor poc-A1 10.10.11.25 6379 2 sentinel down-after-milliseconds poc-A1 3000 sentinel failover-timeout poc-A1 6000 sentinel parallel-syncs poc-A1 1
14:01:05 GMT <michal88> sentinel monitor poc-A1 10.10.11.25 6379 2 sentinel down-after-milliseconds poc-A1 3000 sentinel failover-timeout poc-A1 6000 sentinel parallel-syncs poc-A1 1
14:01:13 GMT <michal88> sentinel monitor poc-A1 10.10.11.25 6379 2
14:01:24 GMT <michal88> sentinel down-after-milliseconds poc-A1 3000
14:01:33 GMT <michal88> sentinel failover-timeout poc-A1 6000
14:01:39 GMT <michal88> sentinel parallel-syncs poc-A1 1
14:01:58 GMT <michal88> when I run the redis-cli debug sleep 60
14:04:33 GMT <michal88> then sentinel detects the failure but doesn't run the new election just waits for master to comeback
14:07:13 GMT <michal88> _Wise_: is you config different to mine for sentinel?
14:16:46 GMT <badboy_> how many sentinels do you have? how many redis nodes?
14:19:18 GMT <michal88> badboy_: I have 3 servers, running both sentinel (26379) and redis (6379)
14:21:36 GMT <michal88> I wonder if there some default setting in redis.conf which is preventing the failover from happening, I changed the protected-mode to NO and have BIND on all ports
14:22:40 GMT <_Wise_> michal88: how many sentinels do you have ?
14:22:52 GMT <_Wise_> (oops)
14:23:14 GMT <michal88> _Wise_:  I have 3 sentinels
14:23:26 GMT <_Wise_> I know O:-)
14:25:41 GMT <_Wise_> michal88: the way I do is to start master and slave without any sentinel active
14:26:37 GMT <_Wise_> then I start the sentinels with the 'sentinel monitor mymaster <ip of master> <port> 2' and 'sentinel down-after-milliseconds mymaster 30001'
14:26:54 GMT <michal88> _Wise_: I did that, I started the master, then created slaveof relations, and then started the sentinel on the Master node. and fallowed with the sentinel on slaves
14:27:19 GMT <_Wise_> I also make sure that the config files are writable
14:27:27 GMT <michal88> for sentinel I use a confing file with the setting as the one above
14:28:22 GMT <_Wise_> but all sentinels are monitoring the same master, right ?
14:28:33 GMT <michal88> yes, shouldn't they?
14:28:36 GMT <_Wise_> yes
14:28:44 GMT <_Wise_> they 'register' to the master
14:28:54 GMT <_Wise_> that's how I see this
14:30:12 GMT <_Wise_> I have to test this with 3.2
14:31:02 GMT <michal88> _Wise_: the funny thing is that I had it working in the past that's why I'm confused as what's wrong with the setup
14:31:03 GMT <_Wise_> michal88: and you specify the '2' at the end of the 'sentinel monitor' line ?
14:31:09 GMT <_Wise_> that's the quorum
14:31:36 GMT <michal88> _Wise_: yes, that's why I run with 3 sentinels
14:31:46 GMT <_Wise_> you should check the logs of the sentinels
14:31:57 GMT <_Wise_> you should see -sdown when the master dies
14:32:05 GMT <_Wise_> (subjectively down)
14:32:15 GMT <_Wise_> +sdown, sorry
14:32:30 GMT <_Wise_> then you should see +odown (objectively down)
14:33:06 GMT <michal88> I see that but what I don't see is this: odown master redis-cluster 172.31.16.68 6379 #quorum 2/2
14:33:13 GMT <michal88> try-failover master redis-cluster 172.31.16.68 6379
14:33:31 GMT <michal88> vote-for-leader 178aaa84a6e7d77fbdacbb8cac24f16e525a0122 2
14:34:00 GMT <michal88> there is different ip as the logs are from an archive(when it was working ;( )
14:35:38 GMT <_Wise_> ok
14:35:56 GMT <_Wise_> are you sute the sentinel can talk to each other on tcp/26379 ?
14:36:08 GMT <_Wise_> you can check this with redis-cli -p 26379 <host>
14:36:43 GMT <_Wise_> from each sentinel
14:45:35 GMT <michal88> _Wise_: connectivity is working, It' trying to test this on 3.0 now
14:46:27 GMT <_Wise_> ack
14:53:43 GMT <michal88> _Wise_: BTW are you minitoring more then one redis DB with your sentinel ? i.e. multiple monitors
15:01:21 GMT <michal88> _Wise_: FYI 3.0.6 it's working fine, it must me something either wrong with default redis config or sentinel on redis 3.2.8
15:03:34 GMT <_Wise_> michal88: the sentinel works at redis instance level, I don't think you can go down to db level
15:29:28 GMT <burn> Morning _Wise_
15:41:44 GMT <_Wise_> yo burn
15:42:37 GMT <_Wise_> I'm located in Belgium :)
15:42:47 GMT <_Wise_> so it's the end of the afternoon for me :)
15:47:36 GMT <michal88> _Wise_: thanks for  your help
15:49:03 GMT <_Wise_> michal88: did not really solve your 3.2 issue
15:50:02 GMT <_Wise_> question : in a Master/Slave setup is there a way to set temp keys to the Master without syncing them to the Slave ?
15:50:22 GMT <michal88> _Wise_: Yes, you are right but at least you have answered my call ;)
15:50:55 GMT <burn> Another question: is the diskless replication feature stable?
15:51:09 GMT <burn> I see a big warning line in the stable config file
16:25:08 GMT <michal88> does anyone know if I can use sentinel to monitor multiple instances of redis i.e on different servers or different ports?
16:26:54 GMT <burn> michal88, we do that with different sentinel instances, yes :-)
16:28:32 GMT <michal88> burn: so I can't have the same instance of sentinel monitoring different redises i.e. different monitors?
16:29:08 GMT <michal88> I was trying to have a cluster of 3 sentinels monitoring our redis estate about 5 diffrent clusters
16:54:03 GMT <gitgud> in redis, can datatypes like set, sorted set, lists etc all be put with an expiry date?
16:54:21 GMT <gitgud> the whole thing, not individual objects inside them
16:55:08 GMT <Habbie> as far as i recall, no
16:56:15 GMT <Habbie> i lied
16:56:17 GMT <Habbie> it works
16:57:37 GMT <Habbie> this is pretty clear from the docs btw
16:57:44 GMT <Habbie> they even have an example with rpush
16:58:16 GMT <gitgud> in here? https://redis.io/commands/rpush
16:58:31 GMT <gitgud> it says nothing about expiry
17:06:25 GMT <Habbie> the 'expire' docs say so
17:06:35 GMT <Habbie> it would be pointless to mention this on -all- other commands
17:08:01 GMT <gitgud> ah i see
17:08:03 GMT <gitgud> ty
18:42:54 GMT <orion> Hi. Does anyone know what's wrong with this command?: 127.0.0.1:6379> SET foo 'bar' 10
18:43:00 GMT <orion> I am getting a syntax error.
18:48:09 GMT <bevilaqua> hi guys.. can someone help me with to undestand a redis-benchmark result?
18:48:51 GMT <bevilaqua> https://gist.github.com/tiagobevilaqua/a8bcddc1f4961e1f5deb588dfdea8492
18:49:33 GMT <bevilaqua> It is a redis-server embedded with a Chef Server of Opscode
18:51:54 GMT <broman> can someone please help me? When I do a redis-cli, it goes fine.. when I try to do this through rails console I get "Error connecting to Redis on 127.0.0.1:6379 (Errno::ECONNREFUSED) subscribe failed" ... how can I troubleshoot this?
19:46:28 GMT <gitgud> hello redis people. is there a way i can tie a hash object's expiry to the expiry of an element in a list/set/sorted set?
19:46:43 GMT <gitgud> or is there a workaround?
19:47:13 GMT <gitgud> i have an hashobject carry data and a reference to that key in a set. i want to make this such a way that when that hashobject expires, i can remove the reference to it or put it back in there
19:47:50 GMT <gitgud> see the thing is i do not need to remove the whole set
19:51:17 GMT <gitgud> oh i see. i could use a sorted set with a timestamp :P
19:54:08 GMT <gitgud> but then in LRU cache systems this would not work