08:34:30 GMT anyone know how to start up a redis/sentinel cluster? 08:37:22 GMT one of my redis-server / redis-sentinels went down, and when i restarted the server, it didnt autostart 08:37:31 GMT so i wanted to start up redis-server, and redis-sentinel 08:37:42 GMT but starting redis-sentinel says this: "sentinel directive while not in sentinel mode" 08:56:32 GMT bjpenn: how are you running redis-server ? 08:59:19 GMT looking at the source, the sentinel should have "redis-sentinel" in the actual process name 08:59:24 GMT or have --sentinel passed to it 09:00:10 GMT http://pastebin.com/XBzZcDbq 09:00:23 GMT the comment is wrong, though 09:00:56 GMT it should be '... or if argv[0] contains "redis-sentinel".' 09:01:00 GMT see strstr(3) 09:38:40 GMT hi 09:38:53 GMT join 12:19:24 GMT hello 12:19:44 GMT PHP don't write in redis with this error : 12:19:48 GMT PHP Warning: Unknown: Failed to write session data (redis). Please verify that the current setting of session.save_path is correct 12:20:22 GMT However redis is installed and configured 12:20:30 GMT the php.ini too 12:20:55 GMT the save_path is correct 12:21:52 GMT come what may the problem? 12:23:52 GMT the problem: PHP 12:25:07 GMT yes but ... :) 13:01:36 GMT In case you want to comment -> https://github.com/antirez/redis/pull/2090 13:11:54 GMT antirez_: I don't have that use case for TTL, so I prefer TTL and TOUCH instead of the NOTOUCH option 13:12:46 GMT soveran: thanks, I wonder if I should merge into 3.2.1, I've the feeling this fixes more bugs than the very rare cases it could create problems 13:13:19 GMT is there a way to only monitor specific commands via redis-cli. I'd like to have the same output as 'monitor' but only see 'ltrim' commands. 13:14:13 GMT antirez_: yes, I think merging it into 3.2.1 is a good idea. 13:15:21 GMT gisli: you can pipe to grep 13:15:25 GMT very time people suggest me semantic versioning something inside me dies 13:15:43 GMT :-) 13:15:51 GMT in a way, this fixes a bug in TTL 13:16:08 GMT so 3.2.1 is fine 13:16:49 GMT yep that's what I think 13:17:01 GMT it's unexpected that non-read stuff touch keys 13:17:46 GMT also expiring is probabilistic so "breaking" is a big term here, unless there are use cases where TTL *it is used* on purpose to refresh the last access time :-) 13:18:08 GMT soveran: wow can't believe I didn't think of that haha! But here's another question, is there a way to exit 'monitor' without exiting the redis-cli 13:18:35 GMT not really 13:19:00 GMT oh, i got a question (unrelated): does it break sentinel setups if i attach independent slaves to a sentinel setup? i.e. i have a 2-instance failover setup and a third (external) redis instance that clones from those 2 13:20:16 GMT minus: you can configure slave-priority 0 13:20:46 GMT minus: but as a general rule, I suggest you create that topology and investigate what happens in each case 13:27:28 GMT slave-priority 0 seems like a good solution; i assumed sentinel would just ignore slaves that it hasn't been configured to monitor; adding that external slave wasn't really intended in the first place 13:27:49 GMT soveran: while you are here I also got a question as well, in the pasta carbonara, do you use the egg raw or you quickly cook it for a few seconds? 13:27:55 GMT thanks for your help 13:28:11 GMT <*> antirez_ kidding 13:28:13 GMT antirez_: I think raw and it's cooked with the pasta 13:28:26 GMT btw while I was kidding... RIGHT ANSWER! :-) 13:28:29 GMT that's the way to go IMHO as well 13:28:59 GMT Dudes we have an hell of a regression in Sentinel 13:29:08 GMT better I move my ass about TTL and fix this too 13:29:14 GMT what is it? 13:29:15 GMT so, fixing with the backward incompatible way 13:29:48 GMT soveran: not sure exactly but master does not get reconfigured properly starting with 3.2.0, I don't understand how this is broken, I only changed connection logic apparently, but instead... I need to check 13:29:51 GMT there are issues 13:30:00 GMT ok 13:30:24 GMT bon courage! 13:32:21 GMT :-) 13:35:09 GMT TTL, TYPE, TOUCH and so forth -> https://github.com/antirez/redis/commit/41d804d9dc48292f61fc1e1efd2241404dc9a6e8 14:14:34 GMT Can a redis key be an integer? 14:15:12 GMT Suppose I want to setnx(userId, 0) where userId is an integer, is that possible? 14:27:49 GMT Also where is my redis.conf on OS X 14:27:58 GMT locate redis.conf 14:28:05 GMT doesn't show up 14:28:14 GMT already tried that 14:28:36 GMT /usr/local/etc/redis.conf 14:28:43 GMT Thanks 14:29:24 GMT Was there 14:48:23 GMT I just want to say I've had Redis available on one project for a while because it was required for Resque. But I just implemented some stuff using just Redis... it's a pleasure to use. Thank you. 20:40:16 GMT in my project I'm storing info about CDs, like the label, the disk type etc, so I'm doing hmset collection:1 label %b type %b 20:41:17 GMT then in some other point I need to get all of the collection IDs in the database (the number after the colon in collection:) 20:42:12 GMT so my question is: should I get all the collection names with scan 0 match collection:* or should I maintain a set in redis and update it every time a collection is added or removed? 20:42:27 GMT and simply get all the values in that set from my program? 20:55:59 GMT King_DuckZ: which programming language are you using? 20:56:56 GMT soveran: c++ 20:57:57 GMT soveran: I think in terms of code they will both translate to one liners, except that if I have to keep a separate set I must make sure it stays up to date 20:58:43 GMT https://bitbucket.org/King_DuckZ/dindexer/src/2a696aa62a71edfd15a0de319efbb09be4a603dd/src/backends/redis/?at=dev <---- redis backend in my project 21:00:40 GMT King_DuckZ: ok, I asked because what you are trying to do is a very common pattern for which there are libraries in other languages 21:02:09 GMT but I'm not sure there's one for C++ 21:02:12 GMT there's this: https://github.com/luca3m/redis3m/blob/master/tests/patterns.cpp 21:02:14 GMT soveran: is it? most of what you see in my link is wrapper code to provide a c++ interface to hiredis, so if you know what those other libraries do, I'll try to provide a similar function in my wrapper code 21:02:45 GMT King_DuckZ: check http://ohm.keyvalue.org 21:03:43 GMT soveran: thanks 21:03:56 GMT the basic idea is that you don't use KEYS or SCAN 21:04:12 GMT instead, you keep a set with the IDs of the keys you want to retrieve later 21:05:53 GMT I see, I had a good feeling then... :) I'll try to implement that in a clean way 21:06:15 GMT I guess at some point you might be able to point people to my wrapper :p 21:09:42 GMT King_DuckZ: I hope so! 21:09:57 GMT King_DuckZ: good luck and feel free to ask 21:10:35 GMT King_DuckZ: by the way, Ohm for Ruby, Lua and Crystal uses the same Lua scripts 21:15:01 GMT https://github.com/soveran/ohm-scripts 22:46:53 GMT <_SLM_> Is there a good MIT-licensed GUI for Redis? 22:48:16 GMT there's redis-desktop-manager. that's the only GUI i know of 22:50:54 GMT <_SLM_> Yeah, saw it. Looked promising until I saw it had a LGPL license 22:51:13 GMT <_SLM_> The info on this site was incorrect, https://redislabs.com/blog/so-youre-looking-for-the-redis-gui which said it was MIT 22:54:58 GMT i never found the need for a (general purpose) GUI for redis 22:55:31 GMT general stuff works from the cli, (business-)specific stuff gets a custom UI 22:56:13 GMT <_SLM_> Makes sense 22:56:22 GMT <_SLM_> I suppose I'll do the same :) 22:56:29 GMT <_SLM_> Thanks for your thoughts 22:56:40 GMT you're welcome