| Time |
Nick |
Message |
| 06:01 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 07:07 |
|
rjackson_isl joined #evergreen |
| 08:07 |
|
stephengwills joined #evergreen |
| 08:22 |
|
stephengwills joined #evergreen |
| 08:34 |
|
mantis1 joined #evergreen |
| 08:34 |
|
mmorgan joined #evergreen |
| 08:54 |
|
Dyrcona joined #evergreen |
| 09:43 |
|
jvwoolf joined #evergreen |
| 09:53 |
|
sandbergja joined #evergreen |
| 09:56 |
|
stephengwills joined #evergreen |
| 10:07 |
Dyrcona |
JBoyer: Dunno if you're really around today, but I've discovered something about my 92 collected auto-renewal notices from last Thursday. They're all for the same user! |
| 10:08 |
JBoyer |
This is the 1 day I am around this week. :) |
| 10:08 |
JBoyer |
And yeah, that does sound promising. |
| 10:08 |
JBoyer |
Anything weird about the user account? |
| 10:10 |
Dyrcona |
Well, they had 92 items checked out that had no renewals left.... :) |
| 10:11 |
JBoyer |
Seems perfectly normal to me, heh. |
| 10:11 |
Dyrcona |
:) |
| 10:11 |
Dyrcona |
They also don't have an email address, so they wouldn't have gotten the notices anyway. |
| 10:13 |
Dyrcona |
This patron also has 2x as many notices as the next patron, who has 46. I wonder if it's a timeout issue or......max stanza size? |
| 10:13 |
* Dyrcona |
checks max stanza size on this utility server. |
| 10:14 |
Dyrcona |
max_stanza_size is set to 2,000,000 |
| 10:16 |
Dyrcona |
I suppose that I could dump the template output from the next patron and assume that this one would be double.... |
| 10:18 |
Dyrcona |
This user also tops the list of courtesy collected courtesy notices from 2 days before. |
| 10:20 |
pinesol |
[evergreen|Bill Erickson] LP1857350 Org selector sorts by display value - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=1bed2a7> |
| 10:20 |
pinesol |
[evergreen|Galen Charlton] LP#1857350: (follow-up) fix lint issue - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=0f46da2> |
| 10:20 |
Dyrcona |
I have a lot of them that failed with the 2-day courtesy notices. |
| 10:21 |
Dyrcona |
Well, more users and failed notices, that is. |
| 10:21 |
|
lstratton joined #evergreen |
| 10:31 |
JBoyer |
I don't suppose those circ ids show up anywhere in your logs since their renewals failed? |
| 10:31 |
JBoyer |
I know you said you didn't think they were complete, but it's really weird that a bunch of failed renewals would knock out an AT runner. |
| 10:34 |
Dyrcona |
Well, he auto-renew events (124) are all OK. It's the notices. There was also a problem with the 2-day courtesy notices, 2 days before. I JabberClient is not connected messages at both times, so I'm operating on the assumption that this user's email notices break something. |
| 10:35 |
Dyrcona |
The 124 is the event id. There are only 92 of them. |
| 10:36 |
Dyrcona |
Did anyone come up with a filter for these that would exclude patrons with no email on file? I seem to recall mmorgan asking about/working on that. |
| 10:38 |
JBoyer |
Not that I know of. Would be helpful though. I can't recall if this is the sort of thing that could easily be done in action_trigger_filter.json or if it would need to be a new validator. |
| 10:39 |
Dyrcona |
Yeah, I'm not sure, either. I'd have to check the environment, but I think it could. |
| 10:40 |
Dyrcona |
...be done in the filter. |
| 10:41 |
* mmorgan |
looks for notes about filters to exclude patrons without email |
| 10:43 |
|
stephengwills joined #evergreen |
| 10:43 |
Dyrcona |
So, the email template output for the patron with 46 items is only 9,769 bytes, and does not have an email address, either. It's not the lack of an email address and it's not the size of the template output. |
| 10:44 |
Dyrcona |
It could still be the amount of data, depending on what is fleshed, or it could be a timeout. |
| 10:45 |
Dyrcona |
I suppose I could look at the trigger code to see if chunking/bundling is present and if not, add it. |
| 10:45 |
Dyrcona |
mmorgan: Thanks! I thought you had done it. |
| 11:32 |
|
khuckins joined #evergreen |
| 11:42 |
mmorgan |
Dyrcona: I located my notes, but realize I never actually got it into production :-( Might have a go at it this afternoon. |
| 11:42 |
|
alynn26 joined #evergreen |
| 12:12 |
Dyrcona |
mmorgan: OK. Thanks, anyway! |
| 12:23 |
|
Christineb joined #evergreen |
| 13:05 |
Dyrcona |
Bleh... CSV is too big for LibreOffice. |
| 13:59 |
Dyrcona |
I thought I had detected a pattern, but it doesn't really hold up. |
| 14:00 |
Dyrcona |
I have noticed that old_due_date appears to not get poplulated, even for successful autorenewals. |
| 14:01 |
Dyrcona |
Most of the autorenewal events that get stuck in collected state, seem to start with 1 patron who has a pile of things that have no renewals remaining, like 50+. |
| 14:02 |
Dyrcona |
Then a bunch that are started after remain collected. |
| 14:05 |
mmorgan |
Dyrcona: could there be some piece of data from item info that's breaking it? I'm always suspicious about punctuation causing problems. |
| 14:08 |
Dyrcona |
mmorgan: Maybe. |
| 14:10 |
Dyrcona |
But, I've got one user whose 85 items renewed and got a status of complete for the notice, then got a collected status 3 weeks later when there were no renewals remaining. I'll check if they're the same copies. |
| 14:11 |
mmorgan |
Dyrcona: I only see old_due_date populated when the autorenewal is unsuccessful. |
| 14:12 |
Dyrcona |
I don't think I've seen it poplulated, yet, but I'm only looking at the events of the users who had the most things to renew. I made a CSV and sorted it by a count. |
| 14:14 |
Dyrcona |
Yeah, the renewal notices that succeeded are for the same copies where it didn't succeed when there were no renewals remaining. |
| 14:18 |
|
lbarry joined #evergreen |
| 14:22 |
Dyrcona |
The autorenewal events work. It seems to be the notices that get hung up. |
| 14:26 |
mmorgan |
Did you say the old_due_date doesn't get populated for these failed autorenewals? |
| 14:28 |
Dyrcona |
I have not seen an autorenewal notice event successful or not where old_due_date is anything other than "" in the user_data. |
| 14:29 |
dbs |
Belated update to my Firefox / MARC Batch Edit struggles: tried running Firefox on Linux with a clean profile and didn't encounter the same problem (of the au.id being replaced with the aou.id and thus not loading a list of containers) |
| 14:29 |
dbs |
So it's Ublock Origin, or Cookie Autodelete, or Zotero, or my browser privacy settings, or ... something |
| 14:31 |
Dyrcona |
Forget it, dbs. It's Firefox.... |
| 14:32 |
* dbs |
ignores the provocation |
| 14:32 |
Dyrcona |
It's a famous movie "quote." |
| 14:35 |
Dyrcona |
Have you disabled Ublock Origin? I use it, too. |
| 14:35 |
Dyrcona |
I use it in Chromium, too, and haven't had any trouble with Evergreen yet. |
| 14:37 |
Dyrcona |
Guess I'll just have to set a new checkout limit of 50. :) |
| 14:45 |
dbs |
Well, I just installed the 8 different add-ons I use in Firefox into a clean FF profile and can't trigger the error |
| 14:46 |
dbs |
(Bitwarden, Cookie Autodelete, EZProxy Redirect Foxified, Firefox Multi-Account Containers, GNOME Shell integration, Grammalecte[fr], uBlock Origin, and Zotero Connector) |
| 14:48 |
Dyrcona |
Of those, I use GNOME Shell Integration and uBlock Origin. |
| 14:50 |
dbs |
was inspecting the stored user preferences under Administration -> Workstation but I don't really see how that would make a difference |
| 14:51 |
dbs |
ah well, hopefully it's just something weird & truly unique with this workstation |
| 14:53 |
|
khuckins joined #evergreen |
| 15:18 |
dbs |
oh man. so I'm using Container Tabs in Firefox, and had been testing in a Work container tab |
| 15:18 |
dbs |
for "fun", I created a new container tab called "Testing" and used that to register a new workstation and test MARC Batch Edit |
| 15:19 |
dbs |
and of course it shows all of my record buckets (also called containers, just to confuse things further with overlapping terminology) |
| 15:21 |
dbs |
so... somehow the workstation inside one particular category of container tabs gets that XHR response wrong unless I use network tools to disable caching |
| 15:21 |
dbs |
and I can't reproduce it even within the same browser (as Firefox containers separate cookies & localstorage, etc, from other containers) |
| 15:22 |
Dyrcona |
So, don't do that? |
| 15:24 |
jeff |
i remain intrigued. it sounds like there isn't a reliable way to reproduce the issue? |
| 15:25 |
jeff |
but the nature of the oddness is... puzzling. |
| 15:25 |
jeff |
it did give me cause to read up on the "raced" bit in firefox dev tools. |
| 15:29 |
|
mantis1 left #evergreen |
| 15:37 |
dbs |
Dyrcona: the problem is, I don't know what "that" is that caused the problem in the first place |
| 15:38 |
Dyrcona |
OK. Fair enough. I was thinking maybe don't use the container tabs. |
| 15:38 |
dbs |
jeff: yeah, I've tried matching up as many configuration bits as I can and can't reproduce it outside of this one container |
| 16:21 |
|
sandbergja joined #evergreen |
| 16:47 |
pinesol |
News from qatests: Failed Installing OpenSRF pre-requisites <http://testing.evergreen-ils.org/~live//archive/2019-12/2019-12-30_16:00:03/test.7.html> |
| 17:06 |
|
mmorgan left #evergreen |
| 17:15 |
pinesol |
[evergreen|Mike Risher] lp1855931 wrap text for wide Angular eg-grid column headers - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=fe8a465> |
| 17:15 |
pinesol |
[evergreen|Galen Charlton] LP#1855931: (follow-up) make grid filter control cells wrap as well - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=9077cbc> |
| 17:17 |
|
berick joined #evergreen |
| 17:48 |
Bmagic |
for small Evergreen patches that introduce a new row into a table, is a pgtap test required? To test that the row exists? |
| 17:49 |
Bmagic |
(Example: AutoRenew patch did not have a pgTAP to test AT hook row) |
| 17:51 |
pinesol |
[evergreen|lfloyd] DOCS: LP 1767378 Work Log documentation - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=8296e8c> |
| 17:51 |
pinesol |
[evergreen|lfloyd] Docs: fixed a spacing issue - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=593a93a> |
| 17:51 |
pinesol |
[evergreen|Jane Sandberg] Docs: LP1767378 follow up: adding manual anchor - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=7ac78d5> |
| 17:51 |
pinesol |
[evergreen|Jane Sandberg] Docs: Fixing asciidoc syntax so fop doesn't complain about staff client admin manual - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=2cd4053> |
| 18:07 |
Dyrcona |
Bmagic: Well, obviously, we don't require a test for that, but adding a test would be highly encouraged, at least by me. |
| 18:08 |
Dyrcona |
Retroactively adding tests for the auto-renew feature would make a decent, bite sized bug. Perhaps I'll open one tomorrow. |
| 18:16 |
Dyrcona |
Anyway, I should have signed out a while ago. Catch you all again, tomorrow! |
| 18:41 |
|
abowling1 joined #evergreen |
| 18:59 |
|
abowling joined #evergreen |
| 19:26 |
|
sandbergja joined #evergreen |
| 20:13 |
|
sandbergja joined #evergreen |
| 21:57 |
|
sandbergja joined #evergreen |
