Evergreen ILS Website

IRC log for #evergreen, 2020-08-05

| 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
00:03 sandbergja_ joined #evergreen
01:37 sandbergja_ joined #evergreen
01:47 sandbergja_ joined #evergreen
02:05 sandbergja_ joined #evergreen
02:41 jvwoolf joined #evergreen
04:47 pinesol News from qatests: Failed Installing OpenSRF pre-requisites <http://testing.evergreen-ils.org/~live//arch​ive/2020-08/2020-08-05_04:00:03/test.7.html>
05:28 agoben joined #evergreen
07:16 rjackson_isl_hom joined #evergreen
07:27 rfrasur joined #evergreen
07:41 collum joined #evergreen
07:54 angelo joined #evergreen
08:07 alynn26 joined #evergreen
08:08 mantis1 joined #evergreen
08:43 mmorgan joined #evergreen
09:06 Dyrcona joined #evergreen
09:15 Bmagic @coffee [someone]
09:15 * pinesol brews and pours a cup of Colombia Granja La Esperanza Geisha, and sends it sliding down the bar to felicia
09:15 Bmagic @coffee [someone]
09:15 * pinesol brews and pours a cup of Rusty's Hawaiian Red Caturra Natural, and sends it sliding down the bar to mmorgan
09:15 Bmagic @coffee [someone]
09:15 * pinesol brews and pours a cup of Kenya Thiriku Top Auction Lot, and sends it sliding down the bar to jeff__
09:15 Bmagic @coffee [someone]
09:15 * pinesol brews and pours a cup of Kenya Nyeri AB Gichatha-ini Signature Series, and sends it sliding down the bar to abneiman
09:16 Bmagic I might be going a little stir crazy these days, isolated at home. I've always got pinesol!
09:16 mmorgan Bmagic: Thank you!
09:17 abneiman Bmagic++
09:17 Bmagic yay! coffee FTW
09:17 Bmagic I'm having some Brazil Rose Diamond myself
09:18 Bmagic from my local roaster: https://www.fretboardcoffee.​com/shop/brazil-rose-diamond
09:18 Bmagic and now business:
09:18 Bmagic Has anyone else wrestled with "cannot write to DB" when renewing via OPAC? Found the error on the DB: "invalid input syntax for type timestamp with time zone: "2020-08-04 23:59:59 <-04>+04""
09:23 Bmagic I'm thinking it's happening in action.push_circ_due_time()
09:24 Bmagic probably stemming from  actor.org_unit_ancestor_setting('lib.timezone'
09:24 csharp maybe a typo in the setting?
09:25 Bmagic select * from actor.org_unit_setting where name='lib.timezone'  0 results
09:26 csharp I would look at activity.log for what the actual call is
09:27 csharp (assuming you segregate your logs like we do)
09:31 Bmagic I've got the logs. The osrfsys log doesn't look unusual.
09:34 csharp hmm
09:34 Bmagic pasting in pastebin
09:40 Bmagic https://pastebin.com/TSQqNEW4
09:41 jvwoolf joined #evergreen
09:45 Bmagic There is more to it than that but I think that's the relevant piece. In case it's not. Here's more: https://pastebin.com/TFusXukB
09:47 Bmagic And the PG error: https://pastebin.com/64W1uXRf
09:52 csharp Bmagic: what does "show timezone;" show?
09:52 mmorgan Odd that the due date is 2020-08-04T00:59:59-0400 rather than 2020-08-04T23:59:59-0400
09:53 Bmagic America/New_York
09:53 csharp is that correct for you?
09:53 Bmagic yep
09:53 * csharp assumed it would be America/Chicago
09:56 Dyrcona rhamby: Won't Fix works for "was valid but isn't any longer." It doesn't fit in with current project plans, though a "Fixed Elsewhere" status could be more what is need.
09:59 csharp Bmagic: https://pastebin.com/cQtyfgPH is the source code for that - looks like it sets whatever the incoming due date is to that date + 1 day - 1 second
09:59 Bmagic right, somehow that code is producing invalid SQL (I think)
10:04 rhamby Dyrcona : I like the idea of a "Fixed Elsewhere" but it may be too sidecase to worry about /shrug I have an irrational instinct to have everything fit the boxes they are in
10:12 csharp Bmagic: this should simulate what that function did with that timestamp:
10:12 csharp select (('2020-08-04T00:59:59-0400' AT TIME ZONE current_setting('timezone'))::DATE + '1 day'::INTERVAL - '1 second'::INTERVAL || ' ' || current_setting('timezone'))::TIMESTAMPTZ;
10:13 csharp 2020-08-04 23:59:59-04 is what I get
10:13 Bmagic yeah, looks sane to me.... what in the world could have caused that DB error....
10:13 Bmagic oh wait - I get an error from your query
10:14 csharp same error?
10:14 Bmagic yep!
10:14 Bmagic select current_setting('timezone');
10:14 Bmagic gives me: "<-04>+04" !!!
10:14 csharp aha!
10:15 Bmagic now we are getting somewhere
10:15 Bmagic csharp++
10:15 Bmagic now.... how is this possible
10:17 Bmagic from postgresql.conf: log_timezone = 'America/New_York'
10:17 berick returns 'America/New_York' when I test
10:17 Bmagic looking at another config, perhaps it should be: log_timezone = 'localtime'
10:17 csharp from the docs, show timezone is equivalent to current_setting('timezone')::text
10:18 berick though I do like the idea of living in '<-04>+04'
10:18 csharp Bmagic: timezone, not log_timezone
10:18 Bmagic ah yes
10:18 Bmagic timezone = 'America/New_York'
10:18 csharp berick: maybe that's the glitch that caused 2020!!!
10:19 csharp <astronaut_with_gun>Always has been</astronaut_with_gun>
10:19 berick heh
10:21 csharp troubleshooting Bmagic's issue is way more fun that what I'm supposed to be doing :-)
10:21 csharp not fun for Bmagic of course
10:22 Bmagic so weird - show timezone; and select current_setting('timezone'); produce the same result. But if I login with the evergreen user, it's different....
10:22 Bmagic csharp: I'm not stressed in the least. I appreciate your hlep!
10:22 Bmagic help
10:22 csharp Bmagic: always happy to help
10:25 Bmagic so - the postgres user gives me the "right" answer to select current_setting('timezone');
10:25 Bmagic the evergreen user gives me the strange <-04>+04
10:25 Bmagic I think I know what to do
10:32 csharp https://stackoverflow.com/questions​/6663765/postgres-default-timezone has some useful information
10:32 Bmagic lol - same article I was readaing
10:33 Bmagic google++
10:33 csharp SELECT * FROM pg_timezone_names where abbrev = '<-04>+04';
10:33 csharp maybe?
10:33 Bmagic 0 rows
10:34 csharp maybe play with ~ or like
10:34 csharp (or visually scan the list on the screen :-) )
10:34 Bmagic SELECT * FROM pg_timezone_names where abbrev ~ '04';
10:34 Bmagic gives results, none of them resemble that weird zone expression
10:48 Bmagic the timezone is set per session. It takes a cue from postgresql.conf timezone setting as default (I think) and for some reason, timezone = 'America/New_York' is the issue
11:03 csharp Bmagic: what happens when you do: SET timezone TO DEFAULT; then select current_setting('timezone');
11:03 Bmagic did that, same issue
11:03 csharp hmm
11:04 csharp show timezone_abbreviations ;
11:04 Bmagic one row: Default
11:04 csharp same here
11:05 Dyrcona Bmagic: Perhaps you should take your questions to #postgresql?
11:05 sandbergja joined #evergreen
11:05 Bmagic I'm thinking I need to change the config to: timezone = 'localtime' and restart
11:05 Dyrcona Bmagic: I'm not so sure of that.
11:05 Bmagic at least, that's my current plan
11:05 Bmagic I'll check with #postgresql
11:07 Bmagic 1000+ Nicks in that channel, wowsers PG rocks
11:09 sandbergja joined #evergreen
11:09 Dyrcona Yes, I often go there first for my PostgreSQL questions.
11:09 Dyrcona When I'm confident it is not an Evergreen thing, that is.
11:10 Dyrcona I'll join you in case you want some solidarity.
11:11 Dyrcona Just like in here, be patient. Also, they tend to get upset with long monologues. Try to put as much on one line as you can and be concise.
11:12 Bmagic right on
11:21 Dyrcona And, maybe no one knows the answer. I'm starting to think it is something in your environment and not so much a PostgreSQL issue.
11:21 berick Bmagic: yay, they finally have a dev release of elasticsearch v7 for Perl -- https://github.com/elastic/elasticsea​rch-perl/releases/tag/version_7.30_1  -- i'll look at that before too long
11:22 Bmagic oh baby
11:41 nfBurton joined #evergreen
11:44 jvwoolf1 joined #evergreen
11:45 nfBurton71 joined #evergreen
11:45 collum joined #evergreen
11:47 Bmagic Dyrcona (and csharp and everyone else) - ALTER USER evergreen reset timezone;
11:47 Bmagic thanks #postgresql
11:52 sandbergja_ joined #evergreen
11:58 sandbergja_ joined #evergreen
12:05 jihpringle joined #evergreen
12:06 rjackson_isl_hom joined #evergreen
12:16 csharp Bmagic++ # interesting
12:30 mrisher joined #evergreen
12:30 mrisher joined #evergreen
12:33 Dyrcona depesz++
12:34 Dyrcona He reviewed the Evergreen schema and did some performance testing once before.
12:34 Dyrcona Bmagic++
12:37 Bmagic I remember that! I was excited to meet someone famous, albiet IRC
12:38 Bmagic again - a video clip comes to mind: https://www.youtube.com/watch?v=v3AucZOsjZ4
12:39 Dyrcona Bmagic: I get very different output with the select * from pg_db_role_setting or \drds than you posted in #postgresql.
12:39 Bmagic whoops, that wasn't in English. Here it is: https://www.youtube.com/watch?v=811z3ktTh-U
12:40 Dyrcona I don't have the path or timezone deliberately set for the evergreen user.
12:40 Dyrcona Yeah, depesz runs the explain site and s a few others.
12:41 Dyrcona 16401 |       0 | {log_lock_waits=1}
12:42 Dyrcona I suspect that you did something, or ran something in your database at some point to set the user path and the timezone.
12:44 Bmagic holy cow - this got me laughing https://www.youtube.com/watch?v=b2cauVuJsJE
12:45 Bmagic Dyrcona: yeah, I'm not really sure how this PG setting got set but I'm glad we found it! I can only imagine it's been plaguing the system
12:46 Dyrcona I'll watch those later when I have my earbuds plugged into the laptop.
12:49 Dyrcona Well, it's in the 0757 upgrade script that you shared a link to in #postgresql. I wonder why I don't have it, unless we missed that update or unset it later.
12:50 Bmagic yeah, not sure either
12:50 Bmagic that was one of two places I found it in the source
12:50 Bmagic looks like it was in wrapped into 2.4 upgrade script
12:51 Bmagic but a fresh install of EG (no upgrade scripts) seems to set it like this: "SET search_path = public, pg_catalog;" - excerpted from 000.english.pg..... script
13:08 Dyrcona I don't think it is necessary to add either to the path, and if your user is evergreen, you don't have to add evergreen, either.
14:40 jvwoolf joined #evergreen
15:00 sandbergja joined #evergreen
15:03 sandbergja_ joined #evergreen
15:04 sandbergja joined #evergreen
15:19 sandbergja joined #evergreen
15:22 sandbergja joined #evergreen
15:32 mantis1 left #evergreen
15:37 jihpringle joined #evergreen
15:47 jvwoolf1 joined #evergreen
15:50 angelo joined #evergreen
15:51 angelo is there anywhere i can find a copy of the EG v3.2 database structure?
15:55 rhamby angelo : have you cloned the source from git?
15:55 rhamby the files are in it or are you looking for something like a diagram?
15:55 angelo a diagram
15:56 angelo i need to compare the differences between 3.2 and 3.4
15:56 rhamby no, that'd be pretty messy.  actor.usr links to dozens of tables easily.
15:56 rhamby a number of tables have a lot of links so you could do a visual representation but to make it meaningful you'd have to break it up into subsections
15:57 angelo im trying to import tsv files from a 3.2 install into a 3.4 install but i think the structure is different
15:57 mmorgan angelo: There's a representation here: http://docs.evergreen-ils.org/3.2_schema/
15:57 rhamby it might be, you can look at the upgrade scripts to see what changed
15:58 Dyrcona angleo: Import into a 3.2 schema and then upgrade after.
15:58 Bmagic rhamby angelo: what about a git diff between the fm_IDL.xml on the two git branches?
15:58 rhamby looking at the upgrade scripts would be faster than manually comparing schemas
15:58 Bmagic oh well - yeah, the root of the reason why is probably better addressed than "how"
15:58 rhamby Bmagic: assuming the changes made it into field mapper :)  but yeah, seems like a valid idea
15:59 Dyrcona Functions change more often than table definitions.
15:59 rhamby Dyrcona's suggestion seems valid to me too
15:59 angelo dyrcona can i use the 3.2 schema on the 3.4 install? or do i need to install a whole new server?
15:59 rhamby yeah, and if I remember this is one of the migration-tools exports so it's the major tables, actor.usr, biblio.record_entry etc....
15:59 Bmagic angelo: yep - setup an EG 3.2 server - do your migration, then upgrade
16:00 Dyrcona Just rebuild the database with the 3.2 setup script, then upgrade the database. You can leave the 3.4 code installed on the server.
16:02 Bmagic rhamby: I just wrote you an email then popped over to IRC and saw you were here.... sorry 'bout that
16:02 Dyrcona Well, I missed a step: reinitialize the database with the 3.2 script, load your data, and then upgrade afterward. If your data is from 3.2 that might be easier.
16:03 angelo dyrcona where can i find that setup script?
16:04 rhamby Bmagic: np, I can type my reply here.  BMagic not at this point but we might move it over to github at some point, for now if you have a patch feel free to send it to me
16:05 Dyrcona angelo: Open-ILS/src/support-scripts/eg_db_config
16:05 Bmagic rhamby: I have one - though I might need to figure out how to squash three local non-consecutive commits and then createa a patch file.....
16:06 Dyrcona If you check out 3.2 or a download a 3.2 tarball, drop the existing database, and then run the eg_db_config from 3.2, you'll get a 3.2 schema installed.
16:06 rhamby Bmagic yeah just send it whenever you can
16:06 Bmagic rhamby: found it git format-patch cc1dde0dd^..6de6d4b06 --stdout > foo.patch   - aint google grand
16:07 angelo yeah i only see 3.5 and 3.4 tarballs on the site
16:07 rhamby I search for stuff all the time, anyone who doesn't ... is either lying or well, probably lying.
16:07 rhamby I guess they could have a very narrow set of job duties ....
16:08 Dyrcona angelo: I'd recommend git over a tarball, but if you copy the link location for the 3.4 or 3.5 tarball and change the numbers, you should get the 3.2 tarball.
16:08 Dyrcona We don't delete them from the server, just hide them.
16:08 angelo got it
16:13 Bmagic rhamby: sent it - So here's the deal. I'm embarking on another collectionHQ quest - and will be making more edits to this repo.... might be nice if this patch were in master. I know, it's selfish.
16:13 Dyrcona Sharing is selfish?
16:13 Dyrcona :)
16:14 Bmagic rhamby: also - I realize that the patch I emailed you created a file called: 655.patch. I put that in my repo as a reminder to submit it to you. lol. You can delete that after the merge
16:14 rhamby Bmagic I have to bail as it's game night but it's top of my inbox so I'll look first thing in the AM
16:14 rhamby ok :)
16:14 Bmagic no worries!
16:14 Bmagic rhamby++
16:14 Dyrcona game_night++
16:14 Bmagic game_night++
16:16 Dyrcona Well, the 655.patch should be easy to remove from the patch file before the merge if it is an all-new file.
16:16 Bmagic it is all new
16:17 Dyrcona Just open the patch in an editor and cut the block for that file.
16:17 Bmagic It's very "matrix" - the patch file I sent him, creates another file in the repo for patching git.... but don't run that one :) just the one I emailed <- rhamby
16:17 Dyrcona git apply won't know the difference, neither will patch, or git am for that matter.
16:18 Bmagic Dyrcona: I considered this just before I clicked send. I wasn't sure enough about the syntax as to where to cut the lines from the patch file, and decided to take this route, explaining it to him here :)
16:19 Dyrcona Well, it's usually pretty obvious where sections for files begin in a patch.
16:19 Bmagic I wasn't 100% and didn't want to break the patch file
16:20 Dyrcona Lines that resemble these:
16:20 Dyrcona diff --git a/Open-ILS/xul/staff_client/ser​ver/OpenILS/symbol_overlay.xul b/Open-ILS/xul/staff_client/ser​ver/OpenILS/symbol_overlay.xul
16:20 Dyrcona index 77b593eca7..33414f2912 100644
16:20 Dyrcona --- a/Open-ILS/xul/staff_client/ser​ver/OpenILS/symbol_overlay.xul
16:20 Dyrcona +++ b/Open-ILS/xul/staff_client/ser​ver/OpenILS/symbol_overlay.xul
16:21 Dyrcona rhamby will figure it out, either way.
16:22 * Dyrcona messes with patches all the time.
16:24 Dyrcona I'm pretty sure that a new file will have "-- a/dev/null" instead of the filename for the "old" version.
16:24 Bmagic cool!
16:24 Dyrcona You edited the patch?
16:24 Bmagic nope, I already sent him the one that includes that new file that he can promptly delete
16:25 Dyrcona :)
16:25 Bmagic it's a patch within a patch..... which makes it a little more fuzzy as to which lines are for the outter patch and which ones are for the inner....
16:25 Bmagic in fact... I wonder if git will handle it...
16:26 Dyrcona Probably.
16:29 Dyrcona Well, catch you tomorrow!
16:36 csharp GPLS has a direct conflict with the hack-a-way dates - not sure how we'll be able to reconcile it
17:13 pinesol [evergreen|Galen Charlton] LP#1889251: make call number affix sort keys read-only - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=ffc05ea>
17:20 mmorgan left #evergreen
17:21 pinesol [evergreen|Bill Erickson] LP1867834 Staffcat clear basket on logout - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=ad66d9e>
17:31 jvwoolf1 left #evergreen
17:43 pinesol [evergreen|Jason Etheridge] LP1774892 stripe elements - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=cfabfa7>
17:43 pinesol [evergreen|Jason Etheridge] LP1774892 stripe elements - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=9aa36c7>
17:43 pinesol [evergreen|Jason Etheridge] LP1774892 stripe elements - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=9aada19>
17:43 pinesol [evergreen|Jason Etheridge] LP1774892 release notes - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=5b2e582>
17:50 sandbergja joined #evergreen
17:58 sandbergja joined #evergreen
17:58 pastebot joined #evergreen
18:01 pinesol News from qatests: Testing Success <http://testing.evergreen-ils.org/~live>
18:15 sandbergja joined #evergreen
19:09 sandbergja joined #evergreen
20:45 jlundgren joined #evergreen
21:48 sandbergja joined #evergreen
21:54 sandbergja joined #evergreen
21:54 sandbergja_ joined #evergreen
22:08 sandbergja joined #evergreen
22:38 sandbergja joined #evergreen
23:49 jvwoolf joined #evergreen

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