| 2013-01-18 00:25:26 utc | ypz | I could not get gist.github.com to work for me, but pastebin works: http://pastebin.com/49hHkxub |
| 2013-01-18 00:25:53 utc | ypz | it is basically the gist sample code you gave me yesterday |
| 2013-01-18 00:26:00 utc | jmettraux | ok |
| 2013-01-18 00:26:54 utc | jmettraux | could you please try with adding :after_connect=>proc{|c| c.reconnect = true} to your Sequel.connect(...) ? |
| 2013-01-18 00:27:24 utc | ypz | sure, one sec |
| 2013-01-18 00:27:42 utc | jmettraux | weird, I thought one had to say adapter "mysql2" for mysql2 to kick in, ... |
| 2013-01-18 00:28:19 utc | jmettraux | I tend to use the connection uris, like in :after_connect=>proc{|c| c.reconnect = true} so I'm not really sure |
| 2013-01-18 00:29:50 utc | ypz | OH, I just switched to mysql from msyql2 |
| 2013-01-18 00:30:19 utc | jmettraux | but your error message was stating Mysql2, so I guess it's kind of automatic... |
| 2013-01-18 00:30:59 utc | ypz | mysql2 doesn't like reconnect: |
| 2013-01-18 00:31:01 utc | ypz | NoMethodError: private method `reconnect=' called for #<Mysql2::Client:0x007feceb0d9e10> (Sequel::DatabaseConnectionError) |
| 2013-01-18 00:31:19 utc | jmettraux | ok, let's google for that |
| 2013-01-18 00:31:48 utc | ypz | with mysql adapter, it works about half of the time |
| 2013-01-18 00:31:53 utc | jmettraux | can't find much |
| 2013-01-18 00:32:06 utc | ypz | without reconnect, mysql almost always give the error |
| 2013-01-18 00:32:21 utc | jmettraux | what happens the other half of the time? |
| 2013-01-18 00:32:48 utc | ypz | same error, lost connection duty query |
| 2013-01-18 00:32:52 utc | ypz | during |
| 2013-01-18 00:32:55 utc | jmettraux | could you please retry with :after_connect=>proc{|c| c.send(:reconnect=, true) } |
| 2013-01-18 00:33:13 utc | jmettraux | (also I lost track: are you testing with or without fork?) |
| 2013-01-18 00:34:16 utc | ypz | with fork |
| 2013-01-18 00:34:24 utc | jmettraux | ok |
| 2013-01-18 00:34:35 utc | jmettraux | what does the send/reconnect solution yield? |
| 2013-01-18 00:37:22 utc | ypz | with send/reconnect, mysql adapter seems to be good, mysql2 still give the same error, though less frequent then without it |
| 2013-01-18 00:37:47 utc | ypz | repeated about 10 times for mysql, not a single error |
| 2013-01-18 00:38:14 utc | jmettraux | thanks |
| 2013-01-18 00:38:31 utc | ypz | with mysql2 adapter, got error 2 out of 6 times |
| 2013-01-18 00:38:35 utc | jmettraux | maybe the reconnect switch is completely different for mysql2 |
| 2013-01-18 00:38:46 utc | jmettraux | we would have to dig its documentation and code |
| 2013-01-18 00:38:58 utc | jmettraux | btw, how do you fork? |
| 2013-01-18 00:39:09 utc | ypz | yea, I'll stay with mysql adapter, not sure what's the difference between the two |
| 2013-01-18 00:39:16 utc | jmettraux | ah sorry, seen |
| 2013-01-18 00:39:36 utc | jmettraux | if you take some time to investigate mysql2, you might find a solution |
| 2013-01-18 00:39:44 utc | jmettraux | and pass it here |
| 2013-01-18 00:40:01 utc | ypz | sure, I'll spend some time on that |
| 2013-01-18 00:40:40 utc | jmettraux | or simply connect after the fork |
| 2013-01-18 00:40:58 utc | jmettraux | or avoid forking and start two workers |
| 2013-01-18 00:41:05 utc | jmettraux | (two ruby processes) |
| 2013-01-18 00:42:49 utc | jmettraux | but the reconnect feels good to have |
| 2013-01-18 00:48:49 utc | ypz | you mean start workers from separate scripts? |
| 2013-01-18 00:49:01 utc | jmettraux | or running the same script twice |
| 2013-01-18 00:49:47 utc | ypz | OK, I'll try that |
| 2013-01-18 00:50:32 utc | jmettraux | added a mention to the reconnect = true trick to the ruote-sequel readme: https://github.com/jmettraux/ruote-sequel#readme |
| 2013-01-18 01:09:29 utc | jmettraux | ok, I have to switch office, no irc there, feel free to use the mailing list or the issue tracker, best regards, thanks for the feedback! |
| 2013-01-18 01:10:11 utc | ypz | thanks |