06:37:42 GMT <mystictot> Hi, I'm trying to delete redis list items using python wrapper method lrem. But After deleting half elements it throws
06:37:42 GMT <mystictot> IndexError: list index out of range
06:37:42 GMT <mystictot> . Here is my my code snippet:              https://pastebin.com/HC11mHPq
06:39:32 GMT <mystictot> And When I Reset the l2 variable. It again delete half and after that throws same error.
07:11:57 GMT <zpojqwfejwfhiunz> can you paste the full traceback somewhere?
07:13:49 GMT <zpojqwfejwfhiunz> I think that the mistake is in changing the size of the list while iterating over it
07:14:49 GMT <zpojqwfejwfhiunz> so halfway through, your list has length n/2 and you're still trying to iterate to n
13:41:25 GMT <AlexC_> Moin
13:41:39 GMT <AlexC_> We are running an AWS Redis cluster - how do we delete all keys across all nodes?
13:43:31 GMT <colo-work> never used cluster, but doesn't FLUSHALL just work?
13:43:56 GMT <colo-work> (or maybe, since cluster doesn't support multiple DBs, just FLUSHDB?)
17:26:05 GMT <axelrtgs> Hello, i was having a bit of a dillemma with Redis in pub/sub usage mode and was wondering if anyone was around to shed some insight
17:27:00 GMT <axelrtgs> i have multiple redis instances distributed in many locations in the world and would like to aggregate all the data from these instances into a single redis instance at our branch office
17:27:46 GMT <axelrtgs> i have seen the clustering mode but that would copy the data from every location in the world to every location in the world which for bandwidth and other reasons would be very impracticle.
17:28:10 GMT <axelrtgs> i was hoping there was a way to essentially copy the data from the remote locations to the central one easily
19:00:14 GMT <zpojqwfejwfhiunz> axelrtgs: do you mean that you don't want to sync all the locations, but just want to have a replica of each location also stored in a central server?
19:04:04 GMT <axelrtgs> yes exactly
19:04:32 GMT <axelrtgs> i dont want all locations to sync with each other just a replica stored in the central location aggregating the pub subs from all the remote locations
19:06:28 GMT <axelrtgs> the caveat is i want to be able to subscribe to the central location and see the messages from all the locations but if i subscribe to the remote location only see the messages local to there
19:07:17 GMT <axelrtgs> zpojqwfejwfhiunz: thanks for responding to my query. i appreciate it. i was worried it would get lost in the join/quit messages
19:09:07 GMT <zpojqwfejwfhiunz> I don't have those messages turned on for display; it's worth finding out how to turn that off in larger channels
19:09:24 GMT <zpojqwfejwfhiunz> can you just make each of the edge nodes masters and have a matching central slave for each?
19:10:08 GMT <zpojqwfejwfhiunz> they're working on multi-master replication but it's still far off AFIK https://www.youtube.com/watch?v=b5UTfy3ftzI
19:10:53 GMT <zpojqwfejwfhiunz> I think you really want full consistency within a limited scope, so you'll have to partition your data so less is synchronized between the edges
20:22:29 GMT <axelrtgs> thanks for the info. finding a solution to this many to one pubsub thing seems to be very complex :(