02:07:46 GMT <diezgote> I to all. I need help about redis pub/subs
02:09:19 GMT <diezgote> hi
02:09:33 GMT <diezgote> anything can help me?
03:43:40 GMT <truthadjustr> If I do SADD x 1, SADD x 2, SADD x 3,...  and using python redis client by r.smembers('x') gave [1,3,2]. Is this order consistent if I do another query later?
03:45:03 GMT <truthadjustr> i also ask similar question for hash via r.hvals('xx'), wither the order will change in a later query (no additions, nor deletions)
09:47:36 GMT <redis906> Has anyone used/considered using lists with lpus/brpop commands for building a pub/sub?
09:48:11 GMT <redis906> I need a pub/sub that does not lose messages like redis pub/sub implementation if no subscribers are listening.
09:48:35 GMT <redis906> The main question I think is the performance
09:50:31 GMT <redis906> Oh, this wouldn't work if I have multiple subscribers for the same list...
09:50:53 GMT <redis906> But in this case I can fan out to list per subscriber
12:19:32 GMT <DysonReturns> good day gents. the KEYS command returns keys set yesterday, but when I query them they are "nil" are they evicted or did i write nil values?
12:19:58 GMT <DysonReturns> if they are evicted I expect them to not even show up when in my wildcard search for keys.
12:20:09 GMT <DysonReturns> please advise. (pretty please :) )
12:23:38 GMT <DysonReturns> I query them using "GET" and they return: (nil)     to be exact. They are string values.  Not sure whether this is an expired key or whether I wrote nothing to it.
12:23:54 GMT <DysonReturns> And i need to know, please.
12:23:56 GMT <minus> if redis-cli returns (nil) it doesn't exist
12:27:22 GMT <DysonReturns> But if KEYS brings up the key, it means it got evicted and not written as nil by my code, right?
12:28:14 GMT <DysonReturns> trying to eliminate the possibility that something is writing empty values to redis. if this is what it looks like when it was evicted, that's fine.
13:38:54 GMT <King_DuckZ> hey all, my c++ hiredis wrapper is taking shape more and more :) https://github.com/KingDuckZ/dindexer/blob/dev/src/backends/redis/incredis.hpp
13:39:37 GMT <King_DuckZ> I'll be profiling and splitting it out of dindexer into a separate github project, but it's very usable already - comments and feedback welcome!
13:48:41 GMT <DysonReturns> congrats, homey. keep it gangster.
13:50:01 GMT <DysonReturns> I didn't get quite the answer I was looking for, but thanks all for existing, regardless.
13:50:03 GMT <DysonReturns> take care.
14:07:05 GMT <minus> King_DuckZ: ugly header guard, why not just #pragma once?
14:10:07 GMT <King_DuckZ> minus: as silly as it may sound, I never took the time to look into the difference... it can be done if it causes no problems anywhere
14:11:43 GMT <minus> it's non-standard (afaik) but works everywhere and is not a pain like header guards
14:20:21 GMT <King_DuckZ> I've got it bound to a shortcut in vim, so <C-\ g> makes it for me :) I don't even think about it anymore
14:21:28 GMT <King_DuckZ> but either way, compilers can recognize the pattern and can tell what I want to do, so I don't think it makes any difference
14:21:54 GMT <King_DuckZ> I'm more concerned about functionality, usability and performance
14:22:08 GMT <minus> fair, if you have a macro
14:22:16 GMT <minus> functionality is the same
14:23:01 GMT <minus> i think i've heard something about pragma once letting the compiler optimize compiling with it somehow, though i'm entirely unsure about that
14:28:57 GMT <King_DuckZ> I mean, runtime performance for the end user, like how much you can squeeze out of your redis server
15:24:13 GMT <crised> Is Firebase DB engine, similar to redis?
15:27:59 GMT <TheDracle> So, I have been having some issues with the general stability of Pub/Sub in a very simple implementation. Basically I make a subscription connection, subscribe to 'polo', and then publish 'marco'. The opposite end subscribes to 'marco' and upon receiving 'macro' publishes 'polo'.
15:29:49 GMT <TheDracle> This works fine on my local environment, but when I push it out to AWS lambda, I'm seeing really strange behavior where it intermittently misses published data.
15:30:16 GMT <TheDracle> I'm wondering if I need to somehow verify the subscription has occurred.
17:41:01 GMT <warik> hello, I am experimenting with Redis Cluster, I was able to setup a 6 nodes cluster (3 master and 3 slaves) I have added a new set of nodes and introduced them via "cluster meet" but I can't find how to rebalance the slots equally between all 4 masters nodes. Any help? thank you
18:53:39 GMT <cos> I'm trying to find documentation on the meaning of the status you get when you run redis-cli "info" on a redis sentinel.
18:53:52 GMT <cos> no luck googling for it so far.  anyone know where it is?
18:54:36 GMT <cos> for example, you get a bunch of lines that look like "master5:name=NAME,status=ok,address=IPADDR:PORT,slaves=2,sentinels=3"
18:55:04 GMT <cos> say I want to know what other status values I might see besides "ok".  Or, I want to know if slaves=2 means there are 2 slaves currently up, or 2 slaves known, even if one of them is down.
18:55:13 GMT <cos> I'm not finding the answers to these sorts of questions
19:27:57 GMT <rmariano> hello
19:28:40 GMT <rmariano> Question, I hada problem of Redis not starting
19:28:54 GMT <rmariano> and the log said I had to enable memory overcommit in the SO
19:28:55 GMT <zetsubou> hello rmariano  :)
19:29:18 GMT <rmariano> My question follows: is this really required? is there another workaround?
19:29:38 GMT <rmariano> zetsubou: hello