Evergreen ILS Website

IRC log for #evergreen, 2018-09-21

| Channels | #evergreen index | Today | | Search | Google Search | Plain-Text | summary | Join Webchat

All times shown according to the server's local time.

Time Nick Message
02:40 beanjammin joined #evergreen
03:04 vibrio12 joined #evergreen
03:56 jeaye17 joined #evergreen
05:27 aldum18 joined #evergreen
05:28 lorimark joined #evergreen
06:23 mos_basik24 joined #evergreen
07:08 Dyrcona joined #evergreen
07:18 Dyrcona joined #evergreen
07:38 bdljohn joined #evergreen
08:07 sforshee4 joined #evergreen
08:14 agoben joined #evergreen
08:15 bos20k joined #evergreen
08:21 juliushaertl4 joined #evergreen
08:28 mllewellyn joined #evergreen
08:29 mllewellyn left #evergreen
08:36 mmorgan joined #evergreen
08:41 kmlussier joined #evergreen
09:06 lsach joined #evergreen
09:12 jvwoolf joined #evergreen
09:22 rlefaive joined #evergreen
09:26 jvwoolf1 joined #evergreen
09:31 yboston joined #evergreen
09:35 nfpl joined #evergreen
09:39 gmcharlt noting that webstaffblocker bug 1745427 now has a pull request
09:39 pinesol Launchpad bug 1745427 in Evergreen 3.1 "Web Client: Serials - Predict New Issues Defaults to Previous Pattern" [High,Confirmed] https://launchpad.net/bugs/1745427
09:41 csharp well, that's not pretty: https://pastebin.com/nhp1gfux
09:42 csharp ^^ this is during a vandelay import from an acq marc file
09:43 gmcharlt csharp: ouch
09:43 gmcharlt what's the other query?
09:44 gmcharlt er, blocking query
09:44 mllewellyn joined #evergreen
09:44 mllewellyn left #evergreen
09:45 mllewellyn1 joined #evergreen
09:47 mllewellyn1 left #evergreen
09:47 csharp doing this forensicly so not sure yet
09:48 csharp looks like it happened more than once that day
09:50 jeff after much head-scratching and looking for something caching org unit settings or eating events, i remembered that a reload of (at least) a perl opensrf service is insufficient to pick up a code change. :P
09:50 Dyrcona csharp: Good luck, I think I've encountered a dead lock only once or twice in my 6+ years of working with Evergreen. Used to happen at least once a month on Horizon.
09:51 mmorgan csharp: just a thought, could the marc file possibly have contained the same record more than once?
09:53 csharp mmorgan: I don't think postgres knows/cares about MARC at the level where a deadlock happens - basically it's a situation where two processes are stuck, each waiting on the other to complete
09:54 csharp more internal to postgres than having to do with the data itself
09:55 csharp Dyrcona: I've seen them before, but usually during upgrades where we're doing multi-threaded database updates
09:55 csharp like, y'know, pingest :-)
09:55 Dyrcona Well, looks like they're both trying to update  acq.lineitem, possibly the same one.
09:56 Dyrcona csharp: Browse ingest will do that, which is why it is not parallelized any more.
09:58 Dyrcona It was sooo much fun on Horizon and Sybase...We'd get a whole cascade of locks to the point where libraries were calling in that the "system was frozen." I wrote a little GUI in Java to list all of the locks, so I could sort them, find the culprit, and kill it. Yeah, I could type Ctrl-k on the line in the table and kill a process with it.
09:59 Dyrcona PostgreSQL has been much, much better in that respect.
10:03 Dyrcona Is anyone looking at bug 1781480? If not, I'm inclined to push it, since it works for us at CW MARS.
10:03 pinesol Launchpad bug 1781480 in Evergreen "Activity metric data isn't always retrieved as expected for searches" [Medium,Confirmed] https://launchpad.net/bugs/1781480
10:05 Dyrcona gmcharlt++ # We're also looking at that pullrequest.
10:07 kmlussier Dyrcona: I just finished looking at it and it works for me too. Go ahead and push it.
10:08 JBoyer cesardv, you around?
10:08 Dyrcona kmlussier: You can push it if you prefer. Please add jlundgren's sign-off if you do.
10:09 csharp https://pastebin.com/Qf4jhDfy - both queries
10:09 csharp updating consecutive (but still different) rows
10:10 Dyrcona csharp: Victim of a page lock.
10:13 Putonlalla19 joined #evergreen
10:16 Putonlalla19 was kicked by jeff: spam
10:25 khuckins joined #evergreen
10:31 pinesol [evergreen|Mike Rylander] LP#1781480: Closures remeber values in subtle ways... - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=a0a885f>
10:31 pinesol [evergreen|Mike Rylander] LP#1781480: Include group owner ancestor badges - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=437e7f8>
10:36 csharp this MARC acq file is just 85 records but is creating a deadlock whenever someone has tried uploading it
10:37 csharp I'm about to try myself
10:43 Dyrcona csharp: Did you load the branch that makes Vandelay run in parallel?
10:44 JBoyer Sounds like this might be a helpful collection of records too, if you're able to share it.
10:46 csharp Dyrcona: I don't think so...
10:46 csharp maybe?
10:46 Dyrcona Lp 1514085
10:46 pinesol Launchpad bug 1514085 in Evergreen "Feature Request: Make Vandelay Asynchronous/Stateless" [Wishlist,Fix released] https://launchpad.net/bugs/1514085
10:46 beanjammin joined #evergreen
10:47 csharp no, I know I didn't
10:47 csharp not on production - we were testing it and I'm not sure where Tiffany left off
10:47 Dyrcona Must be something else, then. Carry on. :)
10:47 csharp thanks for thinking of it
10:48 Dyrcona Might be worth trying with that branch on a test system, though. There's a chance it might resolve your issue.
10:50 cesardv JBoyer: hey! Sorry I was in the zone... what's up?
10:51 JBoyer No problem, I was just curious why bug 1691263 added select-on-focus, since some of our catalogers are unhappy about it.
10:51 pinesol Launchpad bug 1691263 in Evergreen 3.0 "webclient wishlist: wrap long fields in MARC editor" [Medium,Fix released] https://launchpad.net/bugs/1691263
10:51 berick csharp: unrelated to your current issue, do you recall how long your 3.1 SQL upgrade took?  these billing updates are a beast...
10:51 JBoyer It looks like there is reference to a setting but I don't see any way to change it.
10:52 * cesardv looks
10:52 JBoyer berick, mine worked out to roughly an hour per 10 million rows.
10:52 JBoyer (for that one script, that is)
10:54 berick JBoyer: ugh, thanks, that's about what i'm seeing.
10:54 berick may have to break this up into post-upgrade update chunks
10:55 jeff which updates are taking that long?
10:55 JBoyer Careful doing that! I screwed that up because our fine generator started running at one point when I had triggers disabled and so I spent 2-3 more 4 hour sessions fixing things. :/
10:55 csharp berick: 10 hours
10:55 berick jeff: bug 1748924
10:55 pinesol Launchpad bug 1748924 in Evergreen "Enhanced Billing Timestamp Support" [Wishlist,Fix released] https://launchpad.net/bugs/1748924
10:55 csharp berick: gonna do some surgical billing deletions
10:56 berick JBoyer: good one.. i'll keep that in mind
10:56 jeff Do you have any non-stock triggers / etc on the tables in question?
10:56 jeff I don't think that matches the time we experienced at all, but I'm looking for my notes.
10:57 jeff It was a killer if I didn't take care of a non-stock mmpbbt-related trigger, though.
10:57 cesardv JBoyer: it's been a while so my memory is pretty JeffSessions-y lol
10:57 cesardv ...but IIRC the selecOnFocus thing wasn't working at all, and I just added the bit of js needed to make it actually do what it intended
10:57 csharp cesardv++
10:58 csharp @who can't recall?
10:58 pinesol cesardv can't recall.
10:58 csharp perfect
10:58 berick heh
10:59 berick jeff: no custom triggers here, just the mat_summary_* triggers
10:59 JBoyer cesardv, that's ok, I'll throw an LP out there for it to either be finished or removed and people can discuss preferences. For some reason it seems to take more clicks than you'd expect to get it to finally de-select everything.
11:00 cesardv csharp: nice
11:00 JBoyer (and people unfamiliar with select on focus usually learn a lot about it once they accidentally type something once...)
11:01 jeff as long as we disabled mat_summary_upd_tgr, the three updates ran a little under six minutes for a set of about 7.5 million money.billing rows.
11:02 jeff and i believe dbwells did just that in a revised copy of the upgrade script.
11:02 cesardv JBoyer: sounds good man, thanks, if I get a sec today I'll check it out
11:03 berick jeff: crazy, mat_summary_upd_tgr is disabled here too
11:03 csharp jeff: that's very encouragin... oh
11:04 csharp I didn't disable any triggers, not having researched anything about what they do
11:04 berick csharp: it's in modern versions of the 3.1 upgrade script
11:05 JBoyer I thought I disabled most/all of them, but I must not have with those timings. I disabled the *new* one at one point and then bit myself in the foot when that train sailed, or something to that effect.
11:06 csharp ah, then yes, that was disabled
11:06 csharp just verified
11:09 jvwoolf1 left #evergreen
11:09 jeff weird. i'm trying to think of where the overall number of rows might impact the rate of the update in this circumstance. maybe something to do with the indexes, but i don't know.
11:10 jeff of course, it might be another reason entirely.
11:11 jvwoolf1 joined #evergreen
11:14 yboston joined #evergreen
11:22 pinesol [evergreen|Galen Charlton] LP#1745427: account for change in prediction patterns - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=1c3b4a9>
11:25 freebullets26 joined #evergreen
11:25 freebullets26 was kicked by jeff: spam
11:30 pinesol [opensrf|Mike Rylander] LP#1703411: Move OpenSRF XMPP attrs to subelement - <http://git.evergreen-ils.org/?p​=OpenSRF.git;a=commit;h=b44fb86>
11:30 pinesol [opensrf|Bill Erickson] LP#1703411 XMPP opensrf sub-element repairs - <http://git.evergreen-ils.org/?p​=OpenSRF.git;a=commit;h=744f0d1>
11:30 pinesol [opensrf|Bill Erickson] LP#1703411 XMPP opensrf element make check repairs - <http://git.evergreen-ils.org/?p​=OpenSRF.git;a=commit;h=9682d15>
11:36 segy10 joined #evergreen
11:41 segy10 was kicked by jeff: spam
11:54 beanjammin joined #evergreen
11:57 jihpringle joined #evergreen
12:06 dmellado21 joined #evergreen
12:19 berick anyone here purging billing/payment data for aged circulations?
12:21 * mmorgan shakes head "no"
12:26 berick thinking money.aged_billing and money.aged_payment would help speed up this grueling 3.1 SQL upgrade.
12:26 berick move ~2/3 of the data out of the way
12:28 * berick tries to remember of the UI ever loads billing data for aged circs
12:29 mmorgan Cash reports?
12:34 csharp berick: that was my plan as well
12:34 berick csharp: oh, interesting...
12:34 csharp but I don't know what it will affect besides reports
12:35 csharp if the data isn't used anymore, we should probably be purging billings/payments on those anyway (or at least providing a setting somewhere to do so)
12:36 csharp in PINES, we aren't aging circs/holds for a year so it wouldn't affect our users much if it's just reports
12:36 berick mmorgan: looks like it would affect cash reports if you ran reports for data old enough to be linked to an aged circulation
12:37 berick csharp: same here for circs, 1 year
12:37 * mmorgan suspected so.
12:38 berick csharp: would you purge or move to aged tables?
12:38 berick not sure what we'd do, need to ask
12:39 berick maybe a 2-step process, move to aged table during circ-anon, but purge after (say) 3 years
12:40 csharp yeah, proabably age rather than fully purge, then actually purge after a configured retention
12:40 csharp that makes sense to me
12:41 csharp sounds git branch-worthy, but I'm up to my eyeballs today and can't really even look :-/
12:41 berick i'll open a LP
12:41 csharp berick++
12:42 mmorgan berick++
12:47 mkaito_15 joined #evergreen
12:47 mkaito_15 was kicked by jeff: spam
12:51 csharp researching bug 1788223, which is not affecting stock EG, but has broken certain template creation in PINES...
12:51 pinesol Launchpad bug 1788223 in Evergreen "Web client reporter: connection between Circulation (& co.) and Simple Record Extracts Broken" [High,Incomplete] https://launchpad.net/bugs/1788223 - Assigned to Chris Sharp (chrissharp123)
12:51 csharp I've diffed fm_IDL.xml with stock master and they are not different aside from a couple of custom sources we've had in place for a while
12:52 TessaM- joined #evergreen
12:52 csharp any other suggestions for where to look?  the problem is that the SQL builder thinks there's a "simple_record" column on biblio.record_entry
12:55 miker csharp: it's probably the link type on bre.simple_record ... does yours says "has_a" instead of "might_have"?
12:56 TessaM- was kicked by jeff: spam
12:56 * csharp looks
12:58 csharp <link field='simple_record' reltype='might_have' class='rmsr' key='id' map='' />
12:58 csharp nope :-(
12:58 miker mmm :(
13:01 miker csharp: 88bdd77b97a85bd6c41b174faa2e828c34d3a948 I think
13:01 pinesol miker: [evergreen|Galen Charlton] LP#1721807: fix webstaff report templates that have might_have and has_many joins - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=88bdd77>
13:02 csharp ah - yeah, I don't have that applied
13:02 miker csharp: if you're on 3.0.2-ish, you wouldn't have that
13:03 hbrennan joined #evergreen
13:03 csharp yeah that's right
13:04 yboston joined #evergreen
13:18 csharp miker++ # that did it! :-)
13:26 miker cool
14:03 berick csharp: holler if bug 1793802 varies from expectation
14:03 pinesol Launchpad bug 1793802 in Evergreen "Wishlist: Age billing/payment data with circs" [Wishlist,New] https://launchpad.net/bugs/1793802
14:03 * csharp 'll be a'hollerin'
14:05 khuckins_ joined #evergreen
14:05 csharp berick++ # looks good to me!  thanks
14:06 berick cool, thanks csharp
14:23 abneiman berick: can I go ahead and mark bug 1708757 as Fix Released?
14:23 pinesol Launchpad bug 1708757 in Evergreen "Publish Hatch to Chrome web store" [Wishlist,Fix committed] https://launchpad.net/bugs/1708757
14:24 berick abneiman: oh, yes, thanks
14:24 abneiman coo, done
14:24 abneiman *cool
14:26 csharp @who is too coo for schoo?
14:26 pinesol jonadab is too coo for schoo.
14:27 * csharp knows people from his hometown who say "schoo" for "school"
14:27 csharp "Carrollton Hah Schoo"
14:28 berick :)
14:33 * jonadab is actually one of those weird nerds who would've liked school, if only there weren't so many disruptive kids.
14:34 csharp jonadab++
14:34 jonadab Except gym class.  Gym class can die in a fire.
14:36 Dyrcona My 11th grade English teacher: "These are the best years of your life." Me: "If these are the best years of my life, then kill me, now." Turned out she was wrong.
14:39 jnatten joined #evergreen
14:41 jnatten was kicked by jeff: spam
14:41 kmlussier I'm guessing there are very few cases where high school represents the best years of your life.
14:43 mmorgan Dyrcona's teacher had an interesting perspective.
14:44 psprint-mob joined #evergreen
14:45 psprint-mob was kicked by jeff: spam
14:46 jonadab Oh, high school wasn't bad.  Middle school was pretty bad, though.
14:46 jonadab College was fantastic.
14:50 csharp I went to high school with the same people I went to kindergarten with, so not a lot of room for rebranding/reinventing yourself
14:50 csharp it wasn't until college when I felt "cool"
14:51 bshum "It's a miracle these people ever got out of the 20th century."
14:51 csharp s/when/that/
14:51 csharp bshum++
14:51 mmorgan bshum++
14:51 kmlussier csharp: Yeah, same here. I grew up on an island so there really was no opportunity to get away, even during non-school hours.
14:52 rlefaive joined #evergreen
14:57 Dyrcona My High School experience honestly wasn't so bad, but I did have a conversation along those lines with my 11th grade English teacher.
15:14 rlefaive joined #evergreen
15:30 dholland17 joined #evergreen
15:32 jonadab Oh, right:  I tend to forget there are people who didn't move every 3 years when they were kids.
15:32 jonadab I mean, I *know*, but...
15:33 * bshum mostly enjoyed his middle school years
15:33 bshum Except that all the kids were crazy about Star Wars.  Instead of Star Trek.
15:33 Dyrcona :)
15:33 bshum That was disappointing.
15:33 gmcharlt pull requests are now ready for bug 1552778 and bug 1789442
15:33 pinesol Launchpad bug 1552778 in Evergreen 3.1 "Web client check-in "effective date" and check-out "specific due date" should also include times" [Medium,Confirmed] https://launchpad.net/bugs/1552778
15:33 pinesol Launchpad bug 1789442 in Evergreen 3.1 "Web client: When editing an hourly due date the time is automatically changed to 12:00 am" [Medium,Confirmed] https://launchpad.net/bugs/1789442
15:34 Dyrcona gmcharlt expects us to work.... :)
15:34 * Dyrcona is testing bshum's branch for Ubuntu 18 support in Evergreen, and it is going well so far.
15:35 gmcharlt of particular note: that branch establishes a clean_ISO8601 in Evergreen, copied and fixed from cleanse_ISO8601 in OpenSRF; other time-manipulation functions are brought over as well, particularly interval_to_seconds() and seconds_to_interval()
15:35 gmcharlt so... not only do I want you all to work... I want you to work hard! ;)
15:35 Dyrcona :)
15:35 bshum Speaking of Ubuntu 18, should we wait till the .1 to put that through?  It wouldn't be the first time we ported distro support in a future release if we wanted to give more time to test things around
15:36 bshum Though we've also just pushed stuff like that before, and called it "initial support" :D
15:36 bshum With more to come later
15:36 gmcharlt bshum: works for me if it works for y'all; distro support need not be strictly tied to Evergreen .0 releases
15:36 Dyrcona Well, it's not much use without OpenSRF released also. The -RC is Monday, and that's really up to berick.
15:36 bshum That's true, OpenSRF is important
15:36 gmcharlt and on my plate for next week now that the webstafblockers (I think) all have patches at this point
15:36 Dyrcona Anyway, time to run the tests and then kick the tires on the web staff client.
15:37 bshum gmcharlt: That's cool, definitely doesn't hurt to test stuff more thoroughly all around, especially since we're going back to CPAN for some of the dependencies again.  And new perl, etc.
15:38 Dyrcona Tests are failing, but I kind of expected that. :(
15:38 bshum Bah humbug :)
15:38 kmlussier gmcharlt: If nobody else gets to it, I'll try to take a look at webstaffblocker patches over the next couple of days.
15:38 gmcharlt kmlussier++
15:40 Dyrcona Actually, has anyone run make check on master, lately?
15:40 bshum Dyrcona: Err, not for a few days :(
15:40 bshum Well not since last week
15:40 bshum When we were doing the bug week
15:41 Dyrcona Well, it might be a problem with newer Perl versions and something odd that we're doing: Failed test 'use OpenILS::Application::Circ::HoldNotify;'
15:42 Dyrcona Yeahp, that's what it looks like: Unescaped left brace in regex is illegal here in regex; marked by <-- HERE in m/\${ <-- HERE EMAIL_SENDER}/ at /usr/local/share/perl/5.26.1/OpenI​LS/Application/Circ/HoldNotify.pm line 358.
15:42 Dyrcona I think that used to just raise a warning.
15:45 Dyrcona Yeahp, in perl 5.22 it says that construction is deprecated, not illegal.
15:50 spont4e4 joined #evergreen
15:57 berick can we delete HoldNotify yet?
15:58 Dyrcona No idea, but the fix isn't too hard. However, I get some nice warnings including a vulnerability from CGI.pm.
16:00 bshum Dyrcona: fwiw, all live tests pass clean for current master as of this writing on my new ubuntu 16.04 box.  So guess it's a unique issue to 18.04's new perl.
16:01 Dyrcona It's something that was deprecated and then became illegal, i.e. it throws a warning on ubuntu 16.04. bshum: Did you also run make check or just make live-check?
16:02 bshum Dyrcona: I just ran livecheck first
16:02 bshum I can go back and do the other
16:02 bshum make check passes, though I do see all those deprecation warnings you speak of
16:03 Dyrcona live tests are also blowing up, but I'm about to call it a day.
16:08 Dyrcona We're definitely not ready for Ubuntu 18.04 and Perl 5.26.1.
16:20 kmlussier joined #evergreen
16:49 pinesol [evergreen|Dan Wells] LP#1791340 Webstaff: Don't backdate when we're not - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=c09dd9a>
16:49 pinesol [evergreen|Galen Charlton] LP#1791340: expand on comment about backdated checkin times - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=f6cea03>
17:08 mmorgan left #evergreen
17:19 pinesol [evergreen|Mike Rylander] LP#1791335: Retain stat cats on item transfer - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=8cdf97e>
17:44 s5529 joined #evergreen
19:12 repys16 joined #evergreen
19:50 LyndsySimon11 joined #evergreen
22:11 YaknotiS22 joined #evergreen
23:34 Lilium20 joined #evergreen
23:46 geosmin8 joined #evergreen

| Channels | #evergreen index | Today | | Search | Google Search | Plain-Text | summary | Join Webchat