| Time |
Nick |
Message |
| 04:34 |
|
kworstell-isl joined #evergreen |
| 07:31 |
|
BDorsey joined #evergreen |
| 07:36 |
|
kworstell-isl joined #evergreen |
| 07:38 |
|
kworstell-isl joined #evergreen |
| 08:22 |
|
rfrasur joined #evergreen |
| 08:29 |
|
mmorgan joined #evergreen |
| 08:29 |
|
sharpsie joined #evergreen |
| 08:29 |
|
rfrasur joined #evergreen |
| 08:36 |
|
Guest40 joined #evergreen |
| 08:41 |
|
mantis1 joined #evergreen |
| 08:43 |
mantis1 |
We're working with Syndentics on a discovery layer and I'll be adding a javascript tag in our code. They ask for us to put it in our bibliographic template page. Does anyone know what file that is? |
| 08:52 |
|
sandbergja joined #evergreen |
| 08:59 |
|
Dyrcona joined #evergreen |
| 09:08 |
pinesol |
News from commits: LP2034969: Update docs to reference EDI cron jobs <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=d9ec9d3295389ca42b8b0c29b18b6ce40e28cdbd> |
| 09:08 |
pinesol |
News from commits: LP2034969: Add EDI Scripts to Example crontab <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=8f3690d9580bad8d4c1fb9b324811b2287f6dd32> |
| 09:08 |
pinesol |
News from commits: LP2034969: Add EDI Scripts to Makefile.am <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=74d46c4f7b726d611cf0151f8b7ecba9fd7791ad> |
| 09:11 |
sandbergja |
Fun fact (and cause for celebration): we have exactly 100 angular client unit tests in main right now! |
| 09:12 |
sandbergja |
Our safety net is getting stronger! Maybe we can make it 200 before the next big angular or ng-bootstrap upgrade. :-) |
| 09:22 |
Dyrcona |
sandbergja++ |
| 09:39 |
|
kworstell-isl joined #evergreen |
| 09:43 |
berick |
sandbergja++ |
| 10:14 |
Dyrcona |
I don't understand why things that work in a crontab on one Ubuntu 20.04 server don't work on another, unless they actually have different packages installed for "cron." |
| 10:14 |
Dyrcona |
They both set the shell to /bin/bash, so maybe bash isn't actually bash on one of them? |
| 10:16 |
JBoyer |
mantis1, do you mean the record detail page? That would be /openils/var/templates(_bootstrap)/opac/parts/record/summary.tt2 |
| 10:21 |
sharpsie |
Dyrcona: I *think* the default shell is dash on Ubuntu, but I haven't checked that in some time |
| 10:21 |
sharpsie |
but bash wouldn't be aliased to dash |
| 10:22 |
sharpsie |
sh would be aliased to dash |
| 10:22 |
sharpsie |
# ls -l /bin/sh |
| 10:22 |
sharpsie |
lrwxrwxrwx 1 root root 4 Jul 18 2019 /bin/sh -> dash |
| 10:23 |
sharpsie |
that's on a 20.04 server here |
| 10:24 |
Dyrcona |
sharpsie: The crontabs both do SHELL = /bin/bash |
| 10:25 |
mantis1 |
JBoyer it may be I can give it a shot in our test server then get back to you |
| 10:27 |
sharpsie |
Dyrcona: ah |
| 10:27 |
Dyrcona |
Anyway, that's small potatoes at the moment. I tried replacing our 2-day courtesy notices with a combined template for autorenewals on the "bad" vm last night, and it choked on 62,321 notices.I'll have to dig into the logs, but the a/t runner isn't running and its stuck with 52000+ "collected" events. |
| 10:27 |
sharpsie |
is LYING=true set :-) |
| 10:27 |
Dyrcona |
Heh. |
| 10:28 |
sharpsie |
in our case, we're seeing A/T die because ejabberd doesn't like the size of the message |
| 10:28 |
Dyrcona |
I figured the first run would be big, because I have to grab 3 days' worth of autorenewals into 1, but sheesh. |
| 10:29 |
Dyrcona |
sharpsie; Set max_stana_size. |
| 10:29 |
sharpsie |
I've upped the max stanza size to like 10 MB |
| 10:29 |
sharpsie |
yeah |
| 10:29 |
Dyrcona |
I should look. That might be my problem. Maybe one of them is bigger than 10MB? |
| 10:29 |
sharpsie |
the other times A/T has died it was the OOM killer |
| 10:30 |
sharpsie |
and breaking the runs up via granularity and spreading them out timewise worked |
| 10:30 |
Dyrcona |
Check the IRC logs, I've had all kinds of fun with a/t not related to either of those. It just can't handle volume very well. |
| 10:30 |
sharpsie |
also, upping RAM on the VM to crazy levels |
| 10:30 |
sharpsie |
Dyrcona: yeah, I've been spotty over the last 1.5 months due to family concerns |
| 10:31 |
sharpsie |
trying to get my mojo back |
| 10:31 |
Dyrcona |
Understandable. I hope all is going as well as can be for your family. |
| 10:31 |
sharpsie |
thanks |
| 10:32 |
mantis1 |
JBoyer++ that was it! |
| 10:32 |
Dyrcona |
I think some of these overnight batch type things could be better handled by a stand-alone program rather than A/T. |
| 10:32 |
Dyrcona |
I should do a proof of concept for auto-renewals. |
| 10:32 |
sharpsie |
Dyrcona: that would be very interesting to me/PINES |
| 10:32 |
JBoyer |
Glad I could help! |
| 10:33 |
|
jihpringle joined #evergreen |
| 10:36 |
Dyrcona |
I forget what it was that I grepped last time to figure out the max_stanza_size thing. It's not in my history, and I think it has changed over time. |
| 10:41 |
Dyrcona |
Hmm.. maybe this 1 vm is just unhealthy. syslog hasn't rotated in a few days. |
| 10:41 |
sharpsie |
Dyrcona: in osrfsys logs (WARN) it will be something with "large message" I think |
| 10:41 |
Dyrcona |
It's 17GB. |
| 10:42 |
sharpsie |
and in ejabberd logs it will say something about "policy" |
| 10:42 |
mantis1 |
JBoyer: sometimes I have a hard time understanding instruction from vendors |
| 10:42 |
sharpsie |
Dyrcona: oof |
| 10:43 |
Dyrcona |
mantis1: That's OK. The vendors often do not understand their own instructions, either. :) |
| 10:43 |
mantis1 |
LOL |
| 10:43 |
JBoyer |
They also rarely know much about how things work in Eg, and just hope that we can fill in the gaps. :) |
| 10:44 |
sharpsie |
yeah, I was on 3M's speed dial for a long time when I finally took the time to learn about SIP2 |
| 10:44 |
Dyrcona |
sharpsie: Guess. I'll nuke it and build a new one. Log rotation not working is probably related to the cron issues. Everyone now an then, you just get a "bad" vm. |
| 10:44 |
sharpsie |
which was crazy - I was like "this is your standard, right?" |
| 10:44 |
sharpsie |
they're very happy when we do their work for them |
| 10:45 |
Dyrcona |
Hee. "Standards." |
| 10:45 |
sharpsie |
Dyrcona: I had a VM stop running cron recently - restarting cron.service resolved it |
| 10:46 |
sharpsie |
but that's the first time I'd seen that |
| 10:46 |
sharpsie |
@blame Ubuntu |
| 10:46 |
pinesol |
sharpsie: I come to bury Ubuntu, not to praise them. |
| 10:46 |
Dyrcona |
sharpsie: I'll try that before I wipe it out. |
| 10:46 |
Dyrcona |
I ditched Ubuntu on my personal laptop recently, in case you missed me "blogging" about it in the channel. :) |
| 10:46 |
sharpsie |
I created a cron job to update a file once per second |
| 10:46 |
sharpsie |
heh |
| 10:47 |
sharpsie |
I've been very happy on Fedora for many years now |
| 10:47 |
Dyrcona |
I switched to Arch. |
| 10:47 |
sharpsie |
maybe it was once per minute... |
| 10:47 |
sharpsie |
whatever */1 is - probably minute |
| 10:47 |
jeff |
minute. |
| 10:47 |
sharpsie |
jeff: thanks :-) |
| 10:48 |
jeff |
even * * * * * is once per minute on most/all cron implementations. |
| 10:48 |
sharpsie |
Dyrcona: ah - I've only played with Arch in VMs |
| 10:48 |
sharpsie |
jeff: makes sense |
| 10:49 |
sandbergja |
Dyrcona: I switched to Endeavour OS for my personal laptop recently -- really enjoying it so far. It's been fun to dip my toes into the arch world. |
| 10:49 |
Dyrcona |
sharpsie: It's actually easier to install it on hardware than on a vm, because of the UEFI setup. At least that was my experience. I did a test install on a VM before going "live." |
| 10:50 |
Dyrcona |
Right, so /var/log/syslog.1 looks like it was rotated 3 days ago, and it is also 17GB. Something's off on this vm. |
| 10:53 |
Bmagic |
mantis1: probably opac/parts/record/summary.tt2 |
| 10:54 |
Dyrcona |
sandbergja: I hadn't seen Endeavour OS before. It looks cool. |
| 10:55 |
Dyrcona |
Heh. Confusing ld for ls is kind of funny: ld:/var/log/syslog.1: file format not recognized; treating as linker script |
| 10:55 |
Dyrcona |
ld:/var/log/syslog.1:1: syntax error |
| 10:56 |
Bmagic |
mantis1: lol, just made it through the backlog, ignore me |
| 10:58 |
Dyrcona |
I'll make sure I have a backup of the crontab that I'm testing, and then wipe this vm and build a replacement. It's probably not worth figuring out the issues. |
| 11:01 |
|
briank joined #evergreen |
| 11:07 |
jeff |
semi offtopic, but also not: anyone with a library card vendor that they can recommend, I'd be interested. Our requirements are fairly typical, but also we supply a list of card numbers (not a range), so the vendor would need to be able to support that also. |
| 11:08 |
pinesol |
News from commits: LP#1840990: The Mark Damaged and Mark Missing dialogs are missing some <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=22a30e43f45a83dfd72b55be17baf5c1bdd0bdb2> |
| 11:23 |
Dyrcona |
I think I'll make this replacement vm via the command line on the host machine. I think that will work. |
| 11:28 |
berick |
before I start unpacking, anyone know of any Perl code that implements oclc Client Credentials Grant? https://www.oclc.org/developer/api/keys/oauth/client-credentials-grant.en.html |
| 11:30 |
|
Christineb joined #evergreen |
| 11:33 |
|
bgillap joined #evergreen |
| 11:42 |
Dyrcona |
The hoops I have jump through to open a remote file sometimes: /ssh:cwmars mossutil|sudo:opensrfmossutil:/home/opensrf/repos/Evergreen |
| 11:42 |
Dyrcona |
Also, there are way too many useless warnings in the log files. |
| 11:42 |
bgillap |
Security people hate productivity. |
| 11:42 |
bgillap |
IMO |
| 11:43 |
Dyrcona |
bgillap: In this case, the "security people" would be me and Bmagic. |
| 11:43 |
Dyrcona |
But, I think It's cool I can type that as a file path in my editor and actually get to open and edit files remotely. |
| 11:44 |
Dyrcona |
I also just realized that I |
| 11:44 |
Dyrcona |
'm looking at the wrong server.... I usually connect to util, not dev, so..... force of habit. |
| 11:46 |
bgillap |
hahah Well I have emotions about it. I enjoy not being crypto locked and I despise the effort it takes. :) |
| 11:49 |
Dyrcona |
OK! The warnings make more sense now, but they're still useless. |
| 11:51 |
|
bgillap joined #evergreen |
| 12:01 |
|
mantis2 joined #evergreen |
| 12:29 |
Bmagic |
bgillap++ |
| 12:40 |
|
BDorsey joined #evergreen |
| 12:46 |
|
BDorsey joined #evergreen |
| 13:49 |
|
dmoore_doppel joined #evergreen |
| 14:09 |
mantis2 |
can items with a 'display' status be targetted for holds? |
| 14:11 |
Dyrcona |
mantis2: At the moment, no, but there's a Lp bug for that. |
| 14:12 |
Dyrcona |
Lp 1904737 |
| 14:12 |
pinesol |
Launchpad bug 1904737 in Evergreen "Flag/setting for Items with Specific Statuses to Display on Holds Pull List" [Wishlist,Confirmed] https://launchpad.net/bugs/1904737 |
| 14:13 |
Dyrcona |
Feel free to test it and signoff! |
| 14:16 |
mantis2 |
ooo excellent |
| 14:19 |
* jeff |
reviews the comments to see where the "why not use locations for that?" discussion went |
| 14:20 |
* Dyrcona |
doesn't want to get into it. That horse has been talked to death. |
| 14:20 |
Dyrcona |
It's like every other, similar discussion; People gonna do what people gonna do, and not everyone does it the same. |
| 14:22 |
Dyrcona |
Ah, missed an opportunity... shoulda been '...and not everyone gonna do it the same." :) |
| 14:22 |
jeff |
yeah, I just worry sometimes that we're fixing the wrong things. |
| 14:22 |
jeff |
i.e., supporting workarounds instead of making changes that eliminate the need for the workarounds. |
| 14:22 |
Dyrcona |
It's software. There is no "right" or "wrong." |
| 14:22 |
Dyrcona |
Oh, boy..... Here we go.... |
| 14:22 |
jeff |
bigger picture, though. opposite of forest for the trees, maybe? :-) |
| 14:23 |
Dyrcona |
You tell my member libraries' staff that, 'cause I'm frankly done with that conversation. :) |
| 14:24 |
Dyrcona |
They have "workflows" written in wirebound notebooks with yellowed pages.... Yeah, I know.... *Sings the chorus* |
| 14:26 |
Dyrcona |
All of that being said, I like the idea of the flag on the status to show up on holds. It eliminates the "magic" status of some statuses and makes it obvious. |
| 14:27 |
Dyrcona |
Very often good ideas get shot down or delayed because the discussion gets bogged down in unnecessary detail..... |
| 14:27 |
* Dyrcona |
trips off the soapbox he didn't intend to stand on. |
| 14:39 |
Dyrcona |
mantis2: If you're going to test that, it would be helpful if you could assign the bug to yourself in Launchpad. |
| 14:41 |
jeffdavis |
Upshot of the status vs location discussion seems to be "use status for items that should revert to Available on checkin, otherwise use locations"? |
| 14:52 |
mmorgan |
At the risk of tripping over Dyrcona's soapbox, it's the fluidity of the status that's appealing vs. the rigidity of the location. My observation, anyway. |
| 14:55 |
Dyrcona |
So, I figured out why one of my cron jobs was failing. It was trying to talk a database server that doesn't exist. Turns out I was also clever enough to implement the cron job so that if I set PGHOST and PGDATABASE in the crontab it should be fixed. |
| 14:56 |
Dyrcona |
Never figured out why cron/bash seemed totally busted on the one vm. I'm rebuilding it now with twice the RAM: 32GB! |
| 15:22 |
sharpsie |
"soapbox" would be a good server name |
| 15:23 |
Dyrcona |
Right next to sandbox. |
| 15:24 |
sharpsie |
except everything on soapbox is set to maximum verbosity |
| 15:24 |
Dyrcona |
Heh. |
| 15:25 |
mmorgan |
:) |
| 15:25 |
sharpsie |
I used to set up sudo insults (which insults you if you mistype your password) |
| 15:26 |
sharpsie |
and other fun things, but it's been a while since I found that necessary :-) |
| 15:26 |
Dyrcona |
:) |
| 15:33 |
|
mantis2 left #evergreen |
| 16:34 |
|
jihpringle joined #evergreen |
| 17:05 |
|
mmorgan left #evergreen |
| 19:09 |
|
sandbergja left #evergreen |
| 19:14 |
|
dmoore_doppel joined #evergreen |
| 19:24 |
|
dmoore_doppel joined #evergreen |
| 19:43 |
|
dmoore_doppel joined #evergreen |
| 20:22 |
|
sandbergja joined #evergreen |
| 21:33 |
|
sandbergja left #evergreen |
| 22:46 |
|
Guest77 joined #evergreen |
