05:24:54 GMT <shlant> Hi all. I just noticed that when running a command against one of my redis hosts, it returns "(error) READONLY You can't write against a read only slave." BUT the exit code is 0…. that doesn't seem right
05:25:05 GMT <shlant> because of that, my Ansible playbook was showing the command as succeeding
09:50:07 GMT <Guest99069> hi, I am confused that tcpdump got nothing when using redis-benchmark
09:50:47 GMT <Guest99069> but I see the write function call in writeHandler
09:51:09 GMT <Habbie> maybe redis is using a unix socket?
09:52:48 GMT <Guest99069> I did not use -s option
09:53:54 GMT <Habbie> maybe you're dumping the wrong interface
09:54:14 GMT <Guest99069> At first, I used redis command 'monitor', but got nothing. Then I used tcpdump to monitor, still got nothing.
09:54:39 GMT <Habbie> then you're not talking to the same redis instance ;)
09:54:49 GMT <Guest99069> What do you mean by 'dumping the wrong interface'?
09:55:05 GMT <Habbie> you might be running tcpdump on eth0 while redis is talking on lo0
09:56:02 GMT <Guest99069> when I use redis-cli, both redis cmd 'monitor' and tcpdump worked
09:56:56 GMT <Habbie> but redis-benchmark is invisible?
09:57:29 GMT <Guest99069> invisible
09:57:46 GMT <Guest99069> can not see any data from tcpdump
09:59:02 GMT <Guest99069> When I change ip or port when using redis-benchmark, I got 'Writing to socket: Connection refused'. And can see data flowing with tcpdump.
10:22:29 GMT <colo-work> hi there
10:23:16 GMT <colo-work> we have a number of redis servers configured with AOF, and would like to switch those to do periodic RDB dumps instead. is there a recommended migration path for this, that doesn't involve data loss and as little downtime as possible?
11:21:59 GMT <badboy_> colo-work: uhm, just enable it on the fly?
11:23:32 GMT <colo-work> badboy_, ah, I had totally forgotten about setting config options via redis-cli :X
11:23:38 GMT <colo-work> that makes it kinda obvious, i guess
11:23:48 GMT <badboy_> don't forget to persist it to the config file afterwards
11:24:02 GMT <colo-work> yep, sure thing
11:27:19 GMT <colo-work> thanks :)
18:36:42 GMT <gabe__> Hello everyone, I'm having problems starting my redis cluster on my mac, it used to work but not it gives this error: Sorry, the cluster configuration file nodes.conf is already used by a different Redis Cluster node. Please make sure that different nodes use different cluster configuration files.
18:36:58 GMT <gabe__> I posted on stack overflow with more info here: http://stackoverflow.com/questions/42642070/unable-to-start-redis-cluster-servers
18:36:59 GMT <gabe__> Thanks!
18:38:14 GMT <gabe__> *but now it gives this
19:11:58 GMT <BatteryStaple> @gabe__ you can run many redis clusters on a single host. Each instance must bind to a differnt port and utilize a different configuration filename.
19:13:13 GMT <BatteryStaple> @gabe__ chances are something updated a nodes.conf file on your system
19:13:34 GMT <minus> if it crashed before there might still be some lock file, who knows
19:44:05 GMT <gabe__> Thanks @BatteryStaple
19:44:21 GMT <gabe__> looks like a nodes.conf is being generated from whereever I call redis-server from
19:44:27 GMT <gabe__> and that may be the culprit
19:45:11 GMT <gabe__> i had a script that just ran redis-server with an absolute path argument to each of the different redis.conf files
19:45:36 GMT <gabe__> so looks like i actually have to 'cd' into each directory before i can call and start the servers?
22:00:06 GMT <dblessing> Can anyone think of a reason for this error: `ERR unknown command '<86>Dec'`? It seems to only affect Debian 7 and 8 users.
22:00:24 GMT <dblessing> Such a strange string. I have no idea where it came from and why it would have been passed to Redis as a command.