12:36:20 GMT <Ziied> exit
12:36:22 GMT <Ziied> exit
12:36:23 GMT <Ziied> exit
14:15:06 GMT <tgy> what a struggle
18:18:05 GMT <chridal> Hello! I am writing a websocket server in Node, and I am using redis for storing channel/user info so that it can be retrieved by multiple nodes. I am however also consdering to use redis as a message queue.
18:19:05 GMT <chridal> Using PUB/SUB need I be careful of how many channels I sub to? I considered creating one channel for each user, so that when a message for that user appears on the MQ, the server that that user is connected to can pull it off the queue
18:19:55 GMT <chridal> The other option is of course storing which server a user is connected to in redis, and then just pushing it to the channel of that server.
18:20:18 GMT <chridal> Just wondering whether the pub/sub channels are 'expensive', really?
20:21:53 GMT <_xor> Hey guys, quick newbie question :) I'm writing an app that utilizes RabbitMQ + Celery to queue up & execute tasks.  The results of some of those tasks an be binary files (ex: JPG/PNG images, PDFs, etc.).  Those files will never be more than 25mb.
20:22:33 GMT <_xor> Now, the primary storage for the resulting data is any combination of a shared/distributed file system or a rdbms (PostgreSQL most likely in this case).
20:23:00 GMT <wrkrcoop> i dont need a harddrive for redis right?
20:23:37 GMT <_xor> The tasks will be posting at least the result status to Redis.  What I'm contemplating is whether or not to also store the actual resulting file in redis, or if I should have the client(s) pull the data from the file system or rdbms.
20:23:41 GMT <_xor> wrkrcoop: Eh?
20:24:05 GMT <wrkrcoop> _xor: like redis is all in memory so I dont need a harddirve right?
20:24:18 GMT <_xor> Are you actually asking me or are you trying to make a point?
20:24:54 GMT <_xor> If it's the latter, I'm not asking to trying to justify using redis as a primary storage mechanism.
20:25:30 GMT <_xor> I just don't know enough about how redis internally manages memory and how efficient or inefficient it's going to be to store data like that in redis as opposed to requiring each client to directly get it from the file system and/or rdbms.
20:25:34 GMT <wrkrcoop> im going to assume i dont need it
21:11:12 GMT <wrkrcoop> has anyone used aws + redis, i se there’s this thing called elasticcache and they have redis on there …. im wondering if that means the cache size grows