IRC log for #koha, 2016-04-01

All times shown according to UTC.

Time S Nick Message
00:26 mtompset joined #koha
00:26 mtompset Greetings, #koha
00:37 SamEEE joined #koha
00:48 mtompset Any QA folks here?
01:19 drab I'm seeing a lot of "AH01215 No owner passed in or defined!" errors in the log,any idea what those are about?
01:24 rangi not without a line number
01:25 drab at /usr/share/koha/intranet/cgi​-bin/members/ line 98
01:27 rangi are you using the database user, not a real koha user?
01:34 drab I think the guy that's re-adding users etc right now, is, yeah (koha_library)
01:34 rangi yeah he doesn't want to be doing that
01:34 drab I'm seeing those errors for dozens of different borrowersid
01:34 drab oh ok, what should he be doing? I tried to look at guides and stuff like
01:35 rangi it pretty much tells you when you login
01:35 drab but they talk about adding books etc, and don't quite go oevr users/admins and how to set stuff up
01:35 rangi dont use this user :)
01:35 rangi you should make a real user, mark it superlibrarian and use that from then on
01:35 drab right, it does, we just didn't know what else to use to have "superpowers" to add borrowers etc
01:35 drab ah, ok
01:35 drab superlibrarian, that's the bit I missed
01:35 drab thank you
01:37 rangi no worries
01:38 wizzyrea drab: to mark it, you'd go to the user, click the more button, and "set permissions" - then tick superlibrarian
01:38 wizzyrea create a user?
01:39 wizzyrea I thought wahanui used to know how to do that ^.^
01:43 drab yeah, thanks, found a ml post about that and just passed on the info to the guy admin'ing the lib. thanks again
01:46 wizzyrea awesome.
01:46 wahanui That'll be $1 for the awesome jar, wizzyrea
01:46 * wizzyrea gladly pays
01:46 wizzyrea we need more awesome!
01:46 wizzyrea awesome
01:46 wizzyrea awesome
01:46 wizzyrea awesome
01:46 wizzyrea :P
01:47 drab would it be of any use to put up/share an ansible module for koha somewhere? it's not the cleanest and I'm sure some admins will shudder, but at least it takes care of the ubuntu mpm_itk bug etc and will leave you with a working installation
01:47 drab and if so, is there a recommend place for it?
01:48 wizzyrea kohadevbox?
01:48 wahanui kohadevbox is at
01:48 wizzyrea there's ansible stuff there
01:48 wizzyrea in the ansible branch
01:48 wizzyrea drab^
01:50 drab oh cool, I looked on galaxy but didn't see one that was working, didn't know about kohadevbox. thanks
01:51 wizzyrea yep nw
02:04 aleisha_ joined #koha
02:10 aleisha_ been working on my patch for bug 3669 and it got a lot bigger than i expected it to
02:10 huginn Bug[…]w_bug.cgi?id=3669 enhancement, P5 - low, ---, aleishaamohia, Failed QA , Create a template for circ/
02:15 aleisha_ see you all on monday
02:38 irma joined #koha
02:46 wizzyrea later!
03:56 rocio_afk left #koha
03:57 Francesca joined #koha
04:25 Francesca joined #koha
04:44 ccordova joined #koha
04:44 ccordova_ joined #koha
04:45 Francesca joined #koha
04:45 geek_cl joined #koha
04:45 geek_cl Hi guys, eythian is here?
04:56 mtompset @seen eythian
04:56 huginn mtompset: eythian was last seen in #koha 6 hours, 55 minutes, and 26 seconds ago: <eythian> it's slightly terrifying, albeit convenient.
04:57 rangi joined #koha
04:58 geek_cl @seen bad
04:58 huginn geek_cl: I have not seen bad.
04:59 geek_cl is normal a opac with plack load in 13-15 secs.?
04:59 geek_cl anyone experienced something slow with plack ?
05:01 mtompset I'm not familiar with plack, but my thought would be ... is that just the first time?
05:02 cait joined #koha
05:03 geek_cl no, the , and, and ... almost all*.pl
05:04 mtompset Then I don't know. Sorry. Hope you find your answer. Have a great day (24 hour period), #koha cait rangi geek_cl wizzyrea
05:08 geek_cl CGI::param called in list context from package CGI::Compile::ROOT::usr_share_koha_opa​c_cgi_2dbin_opac_opac_2dsearch_2epl line 86, this can lead to vulnerabilities. See the warning in "Fetching the value or values of a single named parameter" at /usr/share/perl5/ line 436.
05:10 Francesca joined #koha
05:14 Francesca joined #koha
05:42 Francesca joined #koha
05:54 kmlussier_ joined #koha
05:56 cdickinson joined #koha
05:57 * cait waves
05:57 magnuse joined #koha
05:58 * magnuse waves
05:59 bshum joined #koha
06:29 jajm hi
06:35 marcelr joined #koha
06:35 marcelr hi #koha
06:35 cait bbiab
06:54 alex_a joined #koha
06:54 reiveune joined #koha
06:54 alex_a bonjour
06:54 reiveune hello
06:54 wahanui kia ora, reiveune
07:02 laurence joined #koha
07:23 sophie_m joined #koha
07:26 wilfrid joined #koha
07:26 paul_p joined #koha
07:32 cait joined #koha
07:32 cait morning #koha
07:35 burdsjm joined #koha
07:41 marcelr hi cait
07:43 sophie_m joined #koha
07:52 cait hi marcelr
08:04 LibraryClaire joined #koha
08:04 LibraryClaire morning #koha
08:16 magnuse good morning LibraryClaire
08:18 LibraryClaire hi magnuse :)
08:45 Francesca joined #koha
09:18 drojf joined #koha
09:18 drojf morning #koha
09:22 LibraryClaire hi drojf :)
09:29 marcelr @later tell pianohacker please look again at bug 14686
09:29 huginn marcelr: The operation succeeded.
09:32 drojf hi LibraryClaire :)
09:38 drojf1 joined #koha
09:49 * cait waves
10:20 drab joined #koha
10:39 francio joined #koha
10:39 francio hello
10:39 wahanui hola, francio
10:40 francio nice to join you
10:40 francio I'm a linux system administrator
10:43 cait be careful - wahanui is a bot :)
10:43 cait I am not, but have to run out for lunch
11:20 francio I see
11:21 francio I'm starting koha installation and migration of a opac catalogue. I'm completely new to the library topic
11:22 francio any suggestion is great
11:23 francio I am starting with a cloud server debian jessie, 8 gb ram 100gb disk, install from git
11:45 meliss joined #koha
11:47 magnuse francio: the recommended way is to install from debian packages, not git
11:47 magnuse packages?
11:47 wahanui rumour has it packages is at
11:50 francio ok
11:50 francio git no good ? unstable?
11:51 francio I go with packages thes
11:51 francio then
11:52 cait git is only for development - packages will be more fun for production :)
11:53 francio :D
11:53 kidclamp joined #koha
11:55 paul_p joined #koha
12:03 sophie_m joined #koha
12:04 drojf1 francio: what system are you migrating from?
12:04 laurence joined #koha
12:08 kidclamp cait!  w00t w00t!
12:09 cait :)
12:09 cait i will come back to the other patches soon too
12:15 NateC joined #koha
12:21 ccordova joined #koha
12:27 geek_cl Hi guys, morning to
12:27 geek_cl we have some fire here _:
12:31 cait geek_cl: have you checked the mysql error message? Illegal mix of collations ?
12:33 Dyrcona joined #koha
12:45 geek_cl cait, yes i will solve that now,
12:45 tcohen joined #koha
12:46 geek_cl but .... when search , appears some internal server error.
12:57 tcohen morning
12:57 tcohen hi geek_cl
12:57 geek_cl morning, hi tcohen ;)
13:05 misilot joined #koha
13:06 JoshB joined #koha
13:12 geek_cl my today fight_
13:13 cait i think this is only a warning - ther is a bug report about it - CGI::param called in list context from package main line 440, this can  lead to vulnerabilities. See the warning in "Fetching the value or  values of a single named parameter" at /usr/share/perl5/ line 436.
13:13 cait bug 16154
13:13 huginn Bug[…]_bug.cgi?id=16154 normal, P5 - low, ---, jonathan.druart, Needs Signoff , Replace CGI->param with CGI->multi_param in list context
13:14 cait the plack related ones i  don't know about
13:17 cait hi tcohen
13:17 cait :)
13:18 geek_cl cait, the problem is, when somebody search, and navigate between page results, sometimes appears "Internal Server Error", randomly, sometime the page 3, others, the 4, after, the page 3 not show that error....
13:19 talljoy joined #koha
13:20 edveal joined #koha
13:26 mtompset joined #koha
13:26 mtompset Greetings, #koha.
13:27 geek_cl o/
13:27 mtompset Greetings, geek_cl
13:27 geek_cl ;)
13:33 * magnuse wanders off to make pizza and wishes #koha all the best
13:34 mario joined #koha
13:45 tcohen hi cait
13:48 alex_a_ joined #koha
13:50 mtompset Greetings, tcohen
13:50 mtompset Greetings, mario alex_a_
13:51 mtompset I just sent an email, but anyone care to look at bug 16185? ;)
13:51 huginn Bug[…]_bug.cgi?id=16185 enhancement, P5 - low, ---, mtompset, Needs Signoff , t/db_dependent/Reports_Guided.t is failing
13:55 amyk joined #koha
13:59 meliss joined #koha
14:00 cait ashimema++
14:04 ashimema :D
14:05 cait one thing less to worry about :)
14:06 cait bug 11087 anyone? ;)
14:06 huginn Bug[…]_bug.cgi?id=11087 enhancement, P5 - low, ---, ketan, CLOSED FIXED, Have the fines tab default to pay fines rather than account
14:06 cait oh
14:06 cait not that on
14:06 cait e
14:06 cait bug 11807
14:06 huginn Bug[…]_bug.cgi?id=11807 enhancement, P5 - low, ---, fcapovilla, Signed Off , Add categorycode conversions to LDAP authentication.
14:06 cait ... no I killed the effect with my typo
14:07 ashimema hgaha
14:07 ashimema just reading that one now cait
14:07 cait :)
14:07 ashimema I presume that was subtly aimed at me anyways ;)
14:11 ashimema I'm on a role cait
14:11 ashimema thankfully this is stuff I understand rather well :)
14:12 ashimema so far at least
14:13 cait :)
14:15 francharb joined #koha
14:27 tcohen is back is back, ashimema is back!
14:28 ashimema for the afternoon at least ;)
14:29 ashimema unfortunately I'm commited to another 16 migrations in the next 2 months.. so I'm rather tied for time still :(
14:29 ashimema but i thought it about time I did some QAing again :)
14:29 ashimema no good being a 'silent partner' ;)
14:36 cait 16!
14:36 cait very glad you took care of those
14:36 cait say hi from time to time ok? even if no time for QA :)
14:37 rocio_afk joined #koha
14:39 cait bye all, have a ncie weekend )
14:39 cait left #koha
14:48 mtompset Oh, and the last one bug 16186
14:48 huginn Bug[…]_bug.cgi?id=16186 major, P5 - low, ---, mtompset, Needs Signoff , t/db_dependent/Circulation_Issuingrule.t is failing
14:54 tcohen mtompset++
14:54 tcohen mtompset++
14:54 tcohen :-D
14:54 tcohen two of them
14:54 wahanui two of them are on jcamins's patches. :)
14:55 mtompset @seen jcamins
14:55 huginn mtompset: jcamins was last seen in #koha 1 week, 0 days, 0 hours, 7 minutes, and 39 seconds ago: <jcamins> Disadvantage of this plan: need to write Perl.
14:55 mtompset :(
14:55 mtompset @later tell jcamins Miss you on the #koha channel. We'll always have 3.6.3 :)
14:55 huginn mtompset: The operation succeeded.
14:58 mtompset It's selfish of me, because I don't like getting the warnings about failed tests. ;)
15:04 alex_a joined #koha
15:14 reiveune bye
15:14 reiveune left #koha
15:15 pianohacker joined #koha
15:21 mtompset tcohen: So the only thing you are waiting on for bug 11592 is a status change and a test plan that triggers the XSLT tweak?
15:21 huginn Bug[…]_bug.cgi?id=11592 major, P3, ---, mtompset, ASSIGNED , opac detail scripts do not respect MARC tag visibility
15:22 mtompset tcohen++ # thanks for the QA passes.
15:27 geek_cl guys, what about this error: cript timed out before returning headers:
15:27 tcohen is the DB accesible?
15:28 tcohen from the server?
15:28 geek_cl tcohen, yes
15:28 geek_cl tcohen, is a remote mysql server, and is connected by a direct cable
15:29 geek_cl all CPU 100%
15:29 geek_cl people running in circles
15:40 edveal joined #koha
15:52 mtompset free -m?
15:52 mtompset df -h
15:53 mtompset paste results :)
15:53 mtompset both servers.
15:54 mtompset I'm wondering if your server is hitting swap thrashing scenarios which slow down everything.
15:55 mtompset BTW, if you have less than 100MB free, I'm pretty sure this is more likely.
15:55 mtompset RAM that is.
15:55 mtompset and no swap.
16:02 laurence left #koha
16:09 pastebot "geek_cl" at pasted "CPU % 100" (39 lines) at
16:09 geek_cl mtompset, ^
16:12 geek_cl mtompset, the image shows a nice view. mostly the 8 cpu at 100%
16:14 pastebot "geek_cl" at pasted "mysql koha data" (23 lines) at
16:21 geek_cl has anyone seen behavior like this?
16:28 geek_cl : 8 cores at 100% , beautiful
16:44 edveal joined #koha
16:53 drojf joined #koha
17:23 cait joined #koha
17:33 ericar joined #koha
18:17 mtompset geek_cl: You still here? I had to go pick up my wife from work.
18:17 geek_cl mtompset, here ;)
18:18 mtompset Your screen shot kind of makes me think my low on memory hypothesis is correct.
18:18 geek_cl i have free memory
18:18 mtompset ps aux | grep apache | wc
18:18 mtompset how much free memory?
18:18 geek_cl 8     103     721
18:19 geek_cl 6459 , of a total of 8005
18:19 mtompset screen shot free -m for me.
18:19 geek_cl perl script consume too muchCPU,
18:19 geek_cl ok i go to screenshot this
18:20 geek_cl mtompset,
18:20 mtompset that's a top.
18:20 mtompset That's not a command line running free -m
18:20 geek_cl
18:21 geek_cl this is free -m :
18:21 mtompset there we go... let's look closer.
18:22 mtompset Okay... next question.
18:23 mtompset Do you ever have 65 people using the OPAC at once?
18:23 geek_cl mtompset, mostly
18:23 geek_cl sometimes less
18:24 geek_cl in staff maybe 20
18:24 mtompset It's my experience that 100MB per apache client.
18:24 mtompset You have 65*100MB.
18:25 mtompset You get 65 people running queries, and apache is going to have a hissy fit.
18:25 geek_cl in fact...
18:25 wahanui in fact is, like, the today situation
18:25 geek_cl :)
18:25 geek_cl yes
18:25 geek_cl i can share some apache confs if you want
18:26 mtompset I'm not an apache expert.
18:26 geek_cl yes, just for do a check list
18:26 geek_cl i use, gzip
18:26 geek_cl deflate
18:26 mtompset But I'm sure there is a way to lower the memory per apache process, such that you can increase the number of people.
18:26 mtompset No, it's okay.
18:26 mtompset I have other bugs to look at.
18:27 geek_cl yes i can imagine that
18:27 mtompset But I am glad that I confirmed my apache is a memory hog hypothesis. :)
18:27 geek_cl the memory is noy the problem. the RAM looks good mostly every time
18:27 geek_cl s/noy/not
18:28 mtompset That screen shot you sent me before had LOTS of apache processes.
18:28 geek_cl in fact, is , many perl process consume all cores at 100%
18:28 mtompset that too.
18:29 geek_cl most of the time, the ram is free
18:29 mtompset I'm sure, but if 65 people click an action all at the same time... your server is getting beaten by apache clients triggering perl scripts.
18:30 geek_cl i see, so where you configure you apache ram process?
18:31 geek_cl mtompset, what mpm worker you use ?
18:31 mtompset itk.
18:31 geek_cl me too
18:31 mtompset what kind of virtualization is this?
18:31 geek_cl VMWARE
18:32 geek_cl 8 cores, 8 GiB RAM
18:32 mtompset who is the provider?
18:32 drab fwiw, if that info is from ps the 100MB is a "lie" as a lot of the 100Mb are actually shared libs
18:32 geek_cl the university customer
18:33 drab also you cannot set the amount of mem to use per process (in prefork) afaik, that would make it not work. what you can limit is the amount of spare child, min/max children and requests per child
18:34 geek_cl drab, in itk ?
18:34 drab I actually don't know itk much, I've never used itk before they asked me to help with a koha instance, the above comment was re mpm_prefork which seems to be connected to itk
18:35 mtompset what OS are you using?
18:36 geek_cl drab, my mpm_prefork.conf look like this :
18:36 geek_cl mtompset, Debian 8.2 64 bit
18:39 pastebot "geek_cl" at pasted "apache2ctl status" (31 lines) at
18:39 drab geek_cl: 150 max request workers seems high to me, but it all really depends on what kind ofload you need to support
18:40 drab also I need to grab lunch, just thuoght I'd comment on the 100MB mem issue to avoid a wild goose chase since as far as I know that's not how it works (ie it's not a straight # requests * child mem = total mem consumed)
18:40 drab which I think is also what you're seeing and why mem is mostly free (and do not, free mem on a server is almost bad/impossible)
18:40 wahanui okay, drab.
18:41 drab a 24/7 ssytem shuold pretty much use up all the memory which you shuold see as "cached"
18:41 geek_cl drab, i will review your comment and perform some extra calcs
18:41 drab the only bad sign of running out of memory is if you hit swap
18:41 geek_cl thanks for your comments
18:42 drab if you don't hit swap then yuo don't have a problem (at least as far as enugh mem goes)
18:42 geek_cl nice lunch ;=
18:42 geek_cl i hace a lower wv.swappiness ;)
18:42 drab what you can do in case there's a leak to prevent a process from grow horribly is to change MaxRequestsPerChild  to say 1000
18:42 geek_cl i have a low vm.swappiness value
18:43 drab that way when it served 1K requests the process will be killed and respawned (which obviously has a cost)
18:43 drab again optimizing for perfs is no joke and not trying to do that here
18:43 drab just trying to share what I've seen, that's all
18:43 drab ttyl
18:43 geek_cl thanks ;=)
18:44 ericar joined #koha
18:45 ericar joined #koha
18:45 huginn New commit(s) kohagit: Bug 16186: t/db_dependent/Circulation_Issuingrule.t is failing <[…]88aceae6ddeede4fa>
18:50 ericar joined #koha
18:52 ericar joined #koha
18:54 ericar joined #koha
18:56 huginn New commit(s) kohagit: Bug 16129: Remove URL::Encode dependency <[…]bd3166caaa4b53f41> / Bug 16138: Restart plack when rotating logfiles <[…]2983536833c69ceea>
19:03 tcohen bye #koha
19:06 huginn New commit(s) kohagit: Bug 16056: Do not crash when searching for an authority if zebra's index is not... <[…]2132c2f1b58b9a981>
19:16 huginn New commit(s) kohagit: Bug 16173: db_structure.t shouldn't have a fixed number of tests <[…]3ef4f1463c9805dc5> / Bug 16185: t/db_dependent/Reports_Guided.t is failing <[…]577f519416f356d99> / Bug 15745: C4::Matcher gets CCL parsing error if term contains ? (question mark) <http://g
19:26 huginn New commit(s) kohagit: Bug 11807: (follow-up) remove date conversions <[…]6a94da8a3da570909> / Bug 11807 : Add support for categorycode conversions when updating an user using... <[…]8016c25cbd507fb44> / Bug 16068: Do not cache overridden prefs <
19:38 huginn New commit(s) kohagit: Bug 10988: Tabs vs Spaces qa failure patch <[…]3cd1c68dfb63f0829> / Bug 10988 - Fix some wording Fix some outdated wording in googleopenidconnect <[…]feacd1c828b6ed93a> / Bug 10988 - Fixes for comments 57 and 58 <
19:48 huginn New commit(s) kohagit: DBRev Bug 10988 - Allow login via Google OAuth2 (OpenID Connect) <[…]7f1a1d373c6036619>
19:53 bag heh heh - I like reading FONT AWESOME!
19:53 bag it’s pretty cool
20:01 geek_cl where i can find the <Location /cgi-bin> config ?
20:02 cdickinson joined #koha
20:08 huginn New commit(s) kohagit: Bug 10612: (QA followup) <[…]1b39452fbb6f9bf4d> / Bug 10612 - Unit tests <[…]26b86e24226146cbc> / Bug 10612: Follow-up - correcting icon to be Font Awesome <[…]18273c85ae736888a
20:09 irma joined #koha
20:32 BobB_Mobile joined #koha
21:03 JoshB left #koha
21:28 huginn New commit(s) kohagit: Bug 7736 - RM followup fixing/removing on .tt <[…]b8035ea5d00654476> / DBRev Bug 7736 - Edifact QUOTE and ORDER functionality <[…]dbea2df975b6d7de9> / Bug 7736 amend to use new style call to Context <
21:38 huginn New commit(s) kohagit: Bug 15630 - Make Edifact module pluggable <[…]698ced25b3a0965d4>
21:40 cait left #koha
22:06 geek_cl joined #koha
22:08 drab I was looking at setting up backups and I've noticed that by default installing from debs a cronjob is created dumping sql and log files backups in /var/spool/koha/$instance_name
22:08 drab is there any reason why the sql is owned by root:koha-library but the logs are root:root ?
22:09 eythian SQL contains library data, but the tar file contains server configs including database passwords, so is more secure.
22:09 eythian That said, the SQL also contains PII so treat it carefully too.
22:10 drab oh, I thoguht it was just logs. yep, fair enough
22:10 drab thanks
22:10 eythian Essentially, it's legitimate for the library to hold a copy of the SQL, less so the config stuff.
22:11 eythian (library as in library staff rather that sysadmins)
22:11 huginn New commit(s) kohagit: Updating the Schema for bugs 7736 and 15630 <[…]5aeb24f3c16d9fd48>
22:16 geek_cl Hi, Can you say what do you think about this strace output, please ?
22:19 geek_cl also, too many PERL *.pm files not founded quickly

| Channels | #koha index | Today | | Search | Google Search | Plain-Text | plain, newest first | summary