2010-01-13T00:13:34 *** wlayton has quit IRC 2010-01-13T00:29:18 *** dbs has quit IRC 2010-01-13T01:46:43 *** mck9 has left #evergreen 2010-01-13T03:05:19 *** eguest309 has joined #Evergreen 2010-01-13T03:17:27 *** gdunbar has quit IRC 2010-01-13T03:42:43 *** frzosima has joined #evergreen 2010-01-13T07:28:26 *** alxp has joined #evergreen 2010-01-13T07:54:52 *** mck9 has joined #evergreen 2010-01-13T07:59:36 *** sfortin has joined #evergreen 2010-01-13T08:13:24 *** natschil has joined #evergreen 2010-01-13T08:32:16 *** gdunbar has joined #evergreen 2010-01-13T09:06:01 *** bshum has joined #evergreen 2010-01-13T09:09:39 *** Meliss has joined #Evergreen 2010-01-13T09:22:42 *** jenny has joined #evergreen 2010-01-13T09:37:27 *** _bott_ has joined #evergreen 2010-01-13T09:47:50 *** atz has joined #evergreen 2010-01-13T09:48:28 *** bshum_ has joined #evergreen 2010-01-13T09:48:41 *** bshum has quit IRC 2010-01-13T09:48:42 *** bshum_ is now known as bshum 2010-01-13T09:54:32 Hello folks. I am having a performance issue (I think) on a certain record which I could use some advice on. 2010-01-13T09:55:17 It is our college yearbook, so we have around 90 volumes and around 350 copies attached. 2010-01-13T09:56:07 The record never loads, so I can only assume it is timing out at some point. 2010-01-13T09:58:23 Ideally I would like to find a way to make the record load faster, but I would also consider extending the timeout interval if that is the simplest solution. I am not sure where to start in the chain, or how to test this circumstance. 2010-01-13T09:59:46 *** grantjohnson has joined #evergreen 2010-01-13T10:03:09 Well, maybe scratch all of that, it seems I am getting some logged errors for this record. Anyone care to hazard a guess on this? 2010-01-13T10:30:07 *** atz has quit IRC 2010-01-13T10:30:47 *** atz has joined #evergreen 2010-01-13T10:36:44 *** brendan_ga_ has quit IRC 2010-01-13T10:47:39 *** grantjohnson has quit IRC 2010-01-13T10:57:54 *** grantjohnson has joined #evergreen 2010-01-13T11:03:55 dbwells: you're saying the Record Summary page in the OPAC doesn't load for the record? 2010-01-13T11:04:40 phasefx: that's correct 2010-01-13T11:05:22 also, if I run: request open-ils.cstore open-ils.cstore.direct.biblio.record_entry.retrieve "189819" in srfsh it eventually times out after 120 seconds 2010-01-13T11:07:25 are the holdings in the MARC? 2010-01-13T11:10:22 yes, though at this point they no longer need to be 2010-01-13T11:14:06 what's the size of the .marc field on the entry? 2010-01-13T11:14:44 what is the easiest way to determine that? 2010-01-13T11:15:25 *** r123 has left #evergreen 2010-01-13T11:16:18 maybe select length(marc) from biblio.record_entry where id = 189819; 2010-01-13T11:16:28 in psql 2010-01-13T11:16:49 yeah, ok 2010-01-13T11:17:03 it says 194858 2010-01-13T11:17:26 what's the size of a record that does load okay in the OPAC? 2010-01-13T11:18:05 Well, the next record has a size of 2462, so needless to say, the first is really large by comparison. 2010-01-13T11:19:17 I'd try removing the holding tags and re-ingesting the record. There's a sql snippet for doing this on the wiki 2010-01-13T11:20:24 *** r123 has joined #evergreen 2010-01-13T11:22:02 Another record with many holdings weighs in at 51892 and loads fine, so there is some threshold between the two. Still a large gap, but not orders of magnitude large. I will try reingesting and see what happens. 2010-01-13T11:24:51 *** mrpeters-isl has quit IRC 2010-01-13T11:25:05 *** mrpeters-isl has joined #evergreen 2010-01-13T11:33:02 *** Dyrcona has joined #Evergreen 2010-01-13T11:35:33 * berick would be curious to know if cstore/postgres log any errors 2010-01-13T11:38:02 *** grantjohnson has quit IRC 2010-01-13T11:38:13 *** jamesrf has joined #evergreen 2010-01-13T11:38:30 and if the ejabberd max_stanza_size is higher than 200K 2010-01-13T11:39:31 *** grantjohnson has joined #evergreen 2010-01-13T11:41:57 I've setup SIP2 for EG following the instructions on the wiki but am getting "Bail out! Read failure in Invalid username" I added the username to the oils_sip.xml file and created it under the group SIP via the staff client, I do have the group SIP under Users with initial editing permission as "group_application.user.sip_client" and then gave the required group permissions via the staff client (EG1.6) any thoughts how I can tell what is passed betw 2010-01-13T11:43:05 work ou's? 2010-01-13T11:43:22 phasefx: well, removing the holdings and reingesting fixed the problem. The reingest itself took 35 seconds, seems a bit excessive. 2010-01-13T11:43:23 moodaepo: you got truncated after "what is passed betw" 2010-01-13T11:43:37 dbwells: see berick's comments above 2010-01-13T11:43:38 I should probably get the latest openncip with the patches from dbs that djfiander last night 2010-01-13T11:44:04 phasefx: thnx "any thoughts how I can tell what is passed between the client and server...logs?" 2010-01-13T11:46:24 * phasefx has never troubleshooted SIP, not sure 2010-01-13T11:46:58 *** grantjohnson has quit IRC 2010-01-13T11:48:13 berick: the ejabberd max_stanza_size is set at exactly 200000 per this page: http://open-ils.org/dokuwiki/doku.php?id=opensrf:1.2:install I am wondering if that is typo, as I see other older docs recommending 2000000. 2010-01-13T11:49:16 dbwells: it should be 2,000,000 2010-01-13T11:49:24 sans commas, of course 2010-01-13T11:49:44 berick++ 2010-01-13T11:50:00 berick: good call, man. I'll update the wiki page too. 2010-01-13T11:50:05 dbwells++ 2010-01-13T11:51:13 moodaepo: should be able to use the logs that are generated 2010-01-13T11:52:53 via syslog 2010-01-13T12:01:32 *** grantjohnson has joined #Evergreen 2010-01-13T12:03:59 atz++ 2010-01-13T12:21:07 *** natschil has quit IRC 2010-01-13T12:21:14 *** natschil has joined #evergreen 2010-01-13T12:25:12 *** dbs has joined #evergreen 2010-01-13T12:31:22 berick: phasefx: expanding the max_stanza_size to 2000000 solved the problem (we had it on one other record as well). Thanks again guys. 2010-01-13T12:39:09 senator++ 2010-01-13T12:41:01 *** natschil has quit IRC 2010-01-13T12:42:31 thanks dbs. i can tell you already that it's broke :-/ but fixes are forthcoming today. 2010-01-13T12:43:41 senator: it would be shocking for something that large not to have broken edges, so as long as the fixes follow, huzzah! 2010-01-13T12:44:46 (my less-than-shiny history of quality control for my own commits also compels me to be lenient towards others) 2010-01-13T12:45:02 hehe 2010-01-13T12:58:51 *** jamesrf has quit IRC 2010-01-13T13:15:07 *** eguest309 has left #Evergreen 2010-01-13T13:34:18 *** jenny has quit IRC 2010-01-13T13:34:38 *** Meliss has quit IRC 2010-01-13T13:44:13 *** B_Bonner has joined #evergreen 2010-01-13T14:15:03 *** jenny has joined #evergreen 2010-01-13T14:17:49 berick: thanks for the reply on list about the Circulate.pm changes. I am happy with keeping my simple solution vs. finding some more optimal solution for all the reasons you give. 2010-01-13T14:18:34 *** jamesrf has joined #evergreen 2010-01-13T14:18:39 berick: would you like me to submit a patch along these lines? 2010-01-13T14:31:21 I'm grabbing upgrade_log id 0133 if no one has it already 2010-01-13T14:32:51 dbwells: that would be great 2010-01-13T14:37:36 make that 0134 2010-01-13T14:56:02 *** djfiander has joined #Evergreen 2010-01-13T15:00:32 *** Meliss has joined #Evergreen 2010-01-13T15:40:14 *** alxp has quit IRC 2010-01-13T15:40:48 Thanks atz and my co-worker figured out why my sip tests were failing. The oils_sip.xml file had two entries for port 6001 which of course is of course not possible. 2010-01-13T15:42:03 yeah, that would be ... bad. 2010-01-13T16:00:28 *** Meliss has quit IRC 2010-01-13T16:07:16 *** bshum has quit IRC 2010-01-13T16:17:04 *** grantjohnson has quit IRC 2010-01-13T16:18:23 *** sfortin has quit IRC 2010-01-13T17:23:40 *** Dyrcona has quit IRC 2010-01-13T17:26:02 *** jenny has left #evergreen 2010-01-13T17:54:55 *** gdunbar has quit IRC 2010-01-13T17:58:00 *** djfiander has left #Evergreen 2010-01-13T17:59:05 *** dbs has quit IRC 2010-01-13T18:04:39 *** r123 has quit IRC 2010-01-13T18:19:35 *** pmplett has joined #evergreen 2010-01-13T19:05:05 *** B_Bonner has quit IRC 2010-01-13T20:09:51 *** jamesrf has quit IRC 2010-01-13T20:21:54 *** atz has quit IRC 2010-01-13T20:48:36 *** wlayton has joined #evergreen 2010-01-13T21:32:01 *** jamesrf has joined #evergreen 2010-01-13T21:57:46 *** jamesrf has quit IRC 2010-01-13T22:18:40 *** brendan_ga has joined #evergreen 2010-01-13T22:27:48 *** dbs has joined #evergreen 2010-01-13T22:49:05 *** jamesrf has joined #evergreen 2010-01-13T22:51:08 *** atz has joined #evergreen 2010-01-13T23:06:23 *** mck9 has left #evergreen