← Previous day | Today | Next day → | Search | Index
All times shown according to UTC.
| Time | Nick | Message |
|---|---|---|
| 20:35 | chris | morning |
| 20:37 | richard | hi |
| 20:37 | chris | hi richard |
| 04:29 | Amit | hi brendan, chris |
| 04:29 | good morning #koha | |
| 04:36 | Jo | Good morning Amite |
| 04:36 | AMit (sorry) | |
| 04:36 | Amit | hi Jo morning |
| 04:36 | Jo | rogue keyboard misbehaving ... |
| 04:44 | Amit | k |
| 04:47 | brendan | hi Amit |
| 04:48 | Jo | hi brendan |
| 04:48 | brendan | hey Jp |
| 04:48 | whoops Jo | |
| 04:48 | Jo | lol |
| 04:48 | brendan | :) |
| 05:01 | Amit | heya brenan |
| 05:01 | sorry brendan | |
| 05:02 | brendan | amit good morning! |
| 05:02 | @wunder new delhi india | |
| 05:02 | munin | brendan: The current temperature in New Delhi, India is 36.0°C (9:30 AM IST on July 06, 2009). Conditions: Haze. Humidity: 37%. Dew Point: 19.0°C. Pressure: 29.53 in 1000 hPa (Steady). |
| 05:07 | Amit | hi munin |
| 05:10 | Jo | off home! good night everybody. |
| 07:00 | Amit | hi nicomo |
| 07:00 | nicomo | hi Amit |
| 07:08 | chris | gah, the rebuild_zebra emails will never stop |
| 07:08 | can i just reply and see man cron, and if you dont understand that, dont try and run your own linux server | |
| 07:37 | Amit | hi kf nahuel |
| 07:37 | nahuel | hi |
| 07:38 | kf | hi nahuel and Amit |
| 07:54 | mason | heh, koha is not your sys-admin :) |
| 07:55 | why not? , and when will this bug be fixed ! | |
| 07:55 | Amit | hi mason |
| 08:00 | hdl_laptop | mason: just do it ;) |
| 08:01 | mason: it "merely" requires not to launh rebuild_zebra over and over again while the previous rebuild_zebra has not finished ;) | |
| 08:02 | chris | lockfile will do that |
| 08:02 | thats not the problem they are having | |
| 08:02 | the problem people are having is understanding what cron is | |
| 08:03 | ie, running rebuild_zebra at all | |
| 08:03 | mason | haaay, i have to ask some-one just to make sure... |
| 08:03 | is the koha-zebraqueue-ctl.sh / zebraqueue_daemon.pl script *not* the prefered way to do zebra updates? | |
| 08:04 | if so - is it because of memory-leaking? | |
| 08:05 | chris | yep |
| 08:05 | mason | what are biblibrians using ? a 1 minute cron of 'rebuild_zebra.pl -z -a -b' |
| 08:05 | or the 'zebraqueue_daemon.pl' daemon? | |
| 08:05 | chris | the memory leak is due to our modules not being safe |
| 08:06 | same problem hurts running under mod_perl | |
| 08:06 | (ie they dont clean up after themselves) | |
| 08:06 | mason | aaah, subtle / complex :/ |
| 08:06 | chris | which is fine when you run them as cgi, or one hit like rebuild_zebra.pl |
| 08:06 | but if it runs forever like mod_perl or the daemon, they leak | |
| 08:07 | im unhappy with the cron solution | |
| 08:07 | mason | gotcha, i recall this being discussed a few months ago |
| 08:07 | chris | but it wont be fixed 3.0.x |
| 08:07 | mason | i guess the other goofy solution is a cron to restart zebraqueue_daemon.pl every 10 mins / 60 mins etc... |
| 08:07 | chris | yeah |
| 08:08 | better solution is to rewrite the daemon and also fix our modules, so we destroy objects properly | |
| 08:08 | mason | ive had the zebraqueue_daemon.pl daemons on my boxes and havent noticed any ballooning processes... |
| 08:08 | chris | but thats harder |
| 08:09 | it takes a while | |
| 08:09 | mason | yeah, ill take a quick tmp fix, until then |
| 08:09 | chris | but if a library circulates or catalogues a lot |
| 08:09 | they can grow fastish | |
| 08:10 | the main problem is, it hasnt been kept up to date | |
| 08:10 | ie when people made changes to rebuild_zebra, they didnt make changes to the zebraqueue-daemon | |
| 08:10 | so its doing slightly different things | |
| 08:10 | maybe harmlessly, maybe not | |
| 08:12 | mason | ok, thanks for that info |
| 08:13 | these circ selenium scripts should provoke some leakage... | |
| 08:44 | chris | mason: http://koha.pastebin.com/m7c124f9b |
| 08:46 | and | |
| 08:46 | http://koha.pastebin.com/m41ad7e76 | |
| 08:46 | proof of concept | |
| 09:37 | mason | ta chris, i'll use that as a start... |
| 09:39 | chris | it works and will only ever have one rebuild_zebra running at a time |
| 09:39 | maybe put a sleep in there | |
| 09:39 | as it stands it will constantly check the zebraqueue table, then run rebuild_zebra if there is anything to be done | |
| 09:40 | for use in production, id probably add a heartbeat to it | |
| 09:40 | so that somethin can monitor it | |
| 09:41 | veki | is it possible to do interlibrary loans in Koha according to ISO ILL protocol? |
| 09:43 | bbl | |
| 09:44 | chris | heh |
| 09:52 | paul_p | good morning everyone ! |
| 10:06 | kf | good morning paul |
| 10:08 | chris | hi kf and paul_p (who is now at lunch) |
| 10:08 | Amit | hi paul_p |
| 10:33 | kf | hi chris :) |
| 11:31 | paul_p | hi chris & kf. (back from lunch ;-) ) |
| 11:54 | |Lupin| | hello ! |
| 11:55 | kf | hi |Lupin| |
| 11:57 | Amit | hi lupin |
| 11:57 | |Lupin| | :) |
| 11:58 | does anybody know how to add a 009 Unimarc field to a record in Perl, please ? | |
| 11:58 | hmm no | |
| 11:58 | actually | |
| 11:58 | how to _create_ a 009 field ? | |
| 11:59 | I tred | |
| 11:59 | MARC::Field->new( '009', '', '', $str); | |
| 11:59 | and also a => $str |
← Previous day | Today | Next day → | Search | Index