01:09:02 GMT hello all, i'm trying to find the best method to do dump redis to a file and import it to another host for triage - im using aws's elasticashe and hoping there is a redis-cli command that can get the job done? seems using --rdb provides me an unknown SYNC command error 06:44:39 GMT is the Multi dimensional indexes mentioned in this article http://redis.io/topics/indexes the same thing as redis' built in geo functions? http://redis.io/commands/geoadd 06:45:14 GMT the article says its not baked into redis but it looks like redis has some similar geospatial searching 12:57:28 GMT hi guys 12:57:29 GMT ** FATAL CONFIG FILE ERROR *** 12:57:29 GMT Reading the configuration file, at line 59 12:57:29 GMT >>> 'sentinel monitor master 10.1.17.175 6379 2' 12:57:30 GMT sentinel directive while not in sentinel mode 12:57:51 GMT i get this when i try to start the daemon with start-stop-daemon 12:58:20 GMT but if i just do exec 2>&1 $BIN /etc/redis/sentinel.conf --sentinel & 12:58:23 GMT it works 12:58:23 GMT ( 12:58:29 GMT why is that? 13:38:36 GMT FINbit: I had to create my own start script for sentinel (Ubuntu 14.04 LTS). 13:41:24 GMT yeah i got where the problem is 13:41:41 GMT i gotta pass some argumentt thats missing to tell redis-server to use --sentinel directive 14:43:08 GMT i have a master slave redis setup. i am in the process of setting up sentinel. i have only two server so can't have the recommended 3 server setup right now. the slave gets promoted to master when master is down. when i bring back the original master it becomes a slave to the new master so all good here. if i take down the new master, the new slave doesn't get promoted to master role. when i bring back the new master service which i 14:44:14 GMT now i have two slaves. the original master reclaims its role only when i restart the service. 14:45:35 GMT any suggestion on what might cause this behavior. 14:52:47 GMT master info: http://pastebin.com/dU7jk7pZ 14:54:15 GMT slave info: http://pastebin.com/s2Br9BXK 14:57:27 GMT slave config: http://pastebin.com/UTvSjKRB 14:59:58 GMT master config: http://pastebin.com/JBEGWLwe 15:05:44 GMT sentinel config: http://pastebin.com/dVLSs8sD 15:16:12 GMT check the sentinel log and what it thinks about the current state 15:16:26 GMT additionally, you will have a hard time with only 2 sentinels. 15:23:13 GMT badboy_: thanks. let me try again to fetch the log info. it rolled over. 15:25:32 GMT badboy_: is it worth pursuing the two server setup?. i can live with few second downtime and an outage if both server go offline. is there any other gotchas i should be worried about?. 15:26:23 GMT well, if any of the sentinels decides differently than the other, nothing will ever advance 15:35:22 GMT badboy_: log info from sentinel server 1: http://pastebin.com/XHtDr8Ke 15:36:13 GMT log info from sentinel server 2: http://pastebin.com/A6tWVav6 15:37:02 GMT i stopped the master. slave got promoted to master role. brought back the original master. it took the slave role but there is something odd. 15:39:42 GMT http://pastebin.com/KbL6bAzz 15:39:58 GMT the new master doesn't see the slave. 15:40:20 GMT the new slave sees the master but says master is down. 15:55:08 GMT badboy_: thanks for taking time. got the point. will go with 3 server based sentinel setup as recommended. 21:42:17 GMT hey, just curious since I can't find it 21:42:28 GMT a result of *0 from SINTER just means empty, correct? 22:01:27 GMT Emn1ty: what's the exact response returned? It should be returning an empty set, not 0 22:07:05 GMT scriptor: *0 22:07:49 GMT I am doing this in the command line, just an fyi 22:17:35 GMT ah, that's what it is 22:17:52 GMT *0 is the number of results, after that it lists the results 22:18:08 GMT so on one that returns 37 items, at the start it displays *37