08:47:19 GMT hi guys... just a quick question... I have a redis master that use 1Gb of RAM, and 2 slaves that use 5GB of RAM... is that normal? there's a way to optimize that? thanks in advance 08:57:01 GMT does redis key need to md5? 09:32:35 GMT That is not a valid sentence, SimpleName 11:53:39 GMT Hi guys 13:00:49 GMT hey, can I set redis-sentinel up to do failbacks as well? 14:14:02 GMT what's a failback? 15:12:28 GMT Let's see... thinking out loud here. I have a heart_rate hash with ids as keys , and temperature hash with ids as keys. And I need to create a CSV file with the id, heart rate and temperature. I guess the best thing would be to just get a unique union the set of keys, then iterate? I don't guess there's a "merge" function in Redis... 15:12:59 GMT My relational-database-background doesn't like this idea, that's why I'm asking. :P 16:20:33 GMT Hey, so I'm trying to SCRIPT LOAD this script and I'm recieving the error `(error) ERR Error compiling script (new function): user_script:2: '=' expected near 'end'`. Neither of those tokens are on line 2, does anyone know what I'm doing wrong? http://ideone.com/VhyeJd 16:32:25 GMT Ah, nevermind. Apparently SCRIPT LOAD doesn't load files 16:47:59 GMT Is anyone here running a ruby app with redis on a FreeBSD machine? There's build errors for hiredis and I'm trying to find a workaround 16:49:21 GMT sorry, I'm on ubuntu. 19:44:11 GMT hi all, I just read the 7 year old email about why redis doesn't have cas 19:44:33 GMT is there a semantic difference between using the WATCH command instead, or did he simply change his mind? 19:45:04 GMT asking since the logic of watch/fail/retry seems very similar to what the CAS logic would be 21:20:32 GMT hi, I am a db newbie. I read about redis as persistent as other dbs and people do use it as a primary store. The only problem is the RAM size. If this is true, can I design like this: for example, With USER, POST, COMMENT data/tables, I store small size data in redis and leave large data in other dbs. All USER data, POST_ID, COMMENT_ID in redis. POST and COMMENT content in other dbs such as mongodb or postgres. Is this a good design or a bad design? 21:26:10 GMT this used to be an active channel but I guess they got tired of answering questions and went elsewhere. 21:27:00 GMT yeah, at least badboy_ used to be quite active 21:27:39 GMT sgo11: splitting stuff into two DBs seems very silly to me because of overhead in management and complexity 21:28:28 GMT you lose the advantage of RDBMS of being able to do joins 21:28:58 GMT and if you have to hit both, redis and db, during a request you also lose the speed advantage of redis 21:29:44 GMT minus, thanks for your reply. I really like redis's speed. but RAM is expensive. For joins, I think I can build indexes in redis although I don't really know how since I haven't read the doc. 21:31:14 GMT don't trade complexity for speed if you can't manage it 21:35:58 GMT for whatever other dbs, I have to do deep learning too, in order to run it correctly with a good performance. Especially, I have to build redis or memcached again on the top of any other dbs. Thus, I am thinking why I can't just learn redis and use it as a persistent db. That's why I had my previous thought. 21:43:43 GMT irc in general finally seems to be taking a downward turn 21:44:13 GMT at least the open source support channels were active back in the 2000s 21:44:13 GMT it's the advent of slack! 21:44:24 GMT yep 21:44:42 GMT and the rise of stackoverflow 21:45:01 GMT and the downfall of forums 21:45:18 GMT It always surprises me that it took so long to come up with the StackOverflow recipe. 21:45:23 GMT SO-style platforms are nice though, if you ask me 21:45:26 GMT I mean, it isn't complicated. 21:45:38 GMT yeah 21:47:14 GMT the older ones tended to have horrible UI