00:46:07 GMT Hey all - I'm trying to think of a way to shoehorn a large data structure into a series of hashes and can't quite figure a good way yet - the dataset looks as follows: https://gist.github.com/DGaffney/6d0ae60e803e9fd9df16522402e5734c. Described in human terms, it's the day-by-day listing of how many people move from one subreddit to another, for all of Reddit, for the entire history of reddit up to January 31 00:46:14 GMT st. So, the data starts in 2005, and is super super small, and gets huge by 2017, so using key as date and subreddit as field, and transitions to that one as value isn't going to work well against hash-max-zipmap-value if I recall correctly? I'm trying to think of how I'd hash all this information and get it nice and even but that's also not leading to any great ideas. Any thoughts? 00:53:05 GMT Perhaps lists with keys of "{DAY}_{SUBREDDIT}" and vals of the flattened transfers (e.g. ["AskReddit", "100", "AskMeAnything", "50" ...])? The whole dataset in JSON representation is around 23-24GB, the machine has 64GB of RAM for use... 12:02:03 GMT hello.. new to redis .. have a question 12:05:43 GMT anyone there? 12:13:05 GMT I use php redis .. trying to push data to queue using "RPUSH" 12:13:17 GMT it suppose to return "length of the queue" 12:13:53 GMT sometimes, when there are high RPS, length of the queue is empty 12:14:25 GMT data getting pushed .. sometime not.. rpush is not throwing exceptions .. 13:09:32 GMT hi everyone. I want to set up a redis cluster and I'm currently reading all the documentation. what is still not clear to me if I really need sentinel. one thing which is confusing is that the redis cluster has a failure detector (as per https://redis.io/topics/cluster-spec) which does automatic failover to slaves while sentinel does provide the same thing "automatic failover" (as per https://redis.io/topics/sentinel). why use sent 13:21:13 GMT brotspinne: it's either cluster or sentinel, you cannot use both 13:22:10 GMT minus: which one should I use? 13:22:29 GMT what do you want it to do? 13:22:37 GMT failover/HA only? sentinel 13:22:44 GMT sharding data? cluster 13:22:47 GMT I want redundancy and sharding 13:23:04 GMT so you need cluster 13:23:19 GMT what is HA? 13:23:25 GMT high availability 13:23:48 GMT hmm I want high availability and sharding. is this possible? 13:24:04 GMT you do get failover with cluster too 13:24:29 GMT so redis cluster is a superset of sentinel? 13:24:49 GMT it works a bit differently, but in principle, yeah 13:24:58 GMT ok this helps a lot 13:29:29 GMT minus: this guy says sentinel isn't for HA either http://stackoverflow.com/a/31143710/616310 I'm a bit confused again 13:32:34 GMT yeah, it just does failover, your client needs to figure out which of the redis machines to connect to (by asking a sentinel maybe, not sure) 14:09:06 GMT minus: do you know if there is a redis python client which supports redis cluster? the recommended one only seems to support sentinel https://github.com/andymccurdy/redis-py 14:10:14 GMT or should I just connect to one of the masters and they will negotiate sharding automatically? 14:11:18 GMT i've never used cluster, no idea 14:12:21 GMT from https://redis.io/topics/cluster-tutorial it seems I should just try to connect to one of the masters and everything else the client should handle automatically but there is so explicit mention about that 14:12:36 GMT there is no explicit mention I mean 14:29:26 GMT you need a smart client to handle it automatically 14:31:11 GMT @badboy_ is there a list of smart clients? 14:31:54 GMT this list doesn't mention cluster support for python-py https://redis.io/clients#python 14:31:57 GMT not that I know of 14:32:21 GMT redis-py I mean 14:33:05 GMT https://github.com/Grokzen/redis-py-cluster 14:34:05 GMT hmm this client is not even mentioned in the official list, I would prefer a well known client 15:30:00 GMT lol 15:30:07 GMT no one added it, that's all 15:30:13 GMT and afaik it's your best shot 15:30:19 GMT it also seems active 19:31:42 GMT Anyone in here successfully configured laravel connected to a redis cluster? 19:31:42 GMT I use AWS Elasticache with cluter-ON on 3 shard with 2 slaves each. 19:31:42 GMT Laravel keeps failing on the MOVED reply of the cluster nodes. 19:31:42 GMT lumen.log: https://plakbord.cloud.nl/p/WOmldaAMkkYpgPD04uoWQPmv 19:49:07 GMT are the talks at the conference this summer going to be available online?