| Time |
Nick |
Message |
| 06:01 |
pinesol |
News from qatests: Failed Running perl live tests <http://testing.evergreen-ils.org/~live//archive/2022-05/2022-05-20_04:00:03/test.49.html> |
| 07:14 |
|
rjackson_isl_hom joined #evergreen |
| 07:34 |
|
collum joined #evergreen |
| 07:44 |
|
RFrasur joined #evergreen |
| 08:44 |
|
mmorgan joined #evergreen |
| 08:57 |
|
jvwoolf joined #evergreen |
| 09:00 |
|
mantis joined #evergreen |
| 09:05 |
|
Dyrcona joined #evergreen |
| 09:06 |
Dyrcona |
Hmm. I'm getting memory warnings in Nagios from two drone servers. Maybe I should test patches in production? |
| 09:09 |
Dyrcona |
Hello, open-ils.storage! |
| 09:10 |
Dyrcona |
Seven storage drones using about 1.5% of RAM each, which isn't a lot, but it adds up, I guess. |
| 09:12 |
Dyrcona |
However, 145 actor drones doesn't help either. :( |
| 09:28 |
Dyrcona |
About 37MB vsz used by the actor drones and listener. Not that much, but I guess it adds up with all of the others. |
| 09:48 |
Dyrcona |
Hmm. My srfsh requests are timing out on my test vm with production data this morning.... |
| 09:50 |
Dyrcona |
It's timing out fleshing call numbers and copies, when I remove the flesh option I get the data back in 0.7 seconds. |
| 09:57 |
Dyrcona |
@monologue |
| 09:57 |
pinesol |
Dyrcona: Your current monologue is at least 8 lines long. |
| 10:12 |
Dyrcona |
Also, the pcrud drone cycles out after hitting about 544KB of rsz, and then a new one takes over that doesn't grow in size. |
| 10:14 |
* gmcharlt |
grabs 1326 |
| 10:14 |
Dyrcona |
Guess this particular VM/db combo is useless for testing at the moment. |
| 10:50 |
|
jonadab joined #evergreen |
| 10:54 |
|
jihpringle joined #evergreen |
| 11:24 |
pinesol |
News from commits: LP#1857910: (follow-up) fix styling of mandatory class and field comboboxes <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=5ac7fc74b7ac34f684c588c97799ac426f1f4035> |
| 11:24 |
pinesol |
News from commits: LP#1857910: (follow-up) enhance IDL for field doc <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=eddc58f4a29afc285931bd87ea3395a2e8a04d0a> |
| 11:24 |
pinesol |
News from commits: LP#1857910: (follow-up) stamp DB update <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=ce88215481014e12b70dd602a5ade5b61c04f8fa> |
| 11:24 |
pinesol |
News from commits: LP#1857910: (follow-up) clean up lint <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=9e39c2a8e1280cb9bfdf602235fcf10e99791b71> |
| 12:28 |
|
collum joined #evergreen |
| 12:52 |
Dyrcona |
So, both patches from bug 1941793 and bug 1941795 cause request open-ils.pcrud open-ils.pcrud.search.bre "AUTHTOKEN" {"id":1}, {"flesh":1,"flesh_fields":{"bre":["call_numbers"]}} to not increase the drone's size at all, even with 100+ requests. |
| 12:52 |
pinesol |
Launchpad bug 1941793 in linux-base (Ubuntu) "Ubuntu linux-base is missing /boot/vmlinu[xz]-* match" [Undecided,New] https://launchpad.net/bugs/1941793 |
| 12:52 |
pinesol |
Launchpad bug 1941795 in xserver-xorg-video-amdgpu (Ubuntu) "Only half of MST tiled display is visible after boot or sleep" [Undecided,New] https://launchpad.net/bugs/1941795 |
| 12:52 |
Dyrcona |
At least on Ubuntu 20.04... |
| 12:52 |
|
collum joined #evergreen |
| 12:52 |
Dyrcona |
Hrm. Wrong bug numbers.... |
| 12:53 |
Dyrcona |
bug 1974193 and bug 1974195 |
| 12:53 |
pinesol |
Launchpad bug 1974193 in OpenSRF "memory leak when C apps respond to client" [High,Confirmed] https://launchpad.net/bugs/1974193 - Assigned to Jason Stephenson (jstephenson) |
| 12:53 |
pinesol |
Launchpad bug 1974195 in Evergreen 3.8 "memory leak when doing fleshed Fieldmapper searches from C DB services" [High,Confirmed] https://launchpad.net/bugs/1974195 |
| 12:56 |
Dyrcona |
Hm. Spoke too soon. Missed the second drone that got started up. |
| 13:47 |
jmurray-isl |
Does anyone have any specific recommendations for sending nginx logs to a central rsyslog over TCP? In my research, it seems that nginx's built in syslog function may only work over UDP, not TCP, and recommendations are to output to a named pipe and send that to the central logger. Keeping track of that just seems inelegant, however. I'd welcome any recommendations beyond that. |
| 13:47 |
Dyrcona |
Yeah, the flesh of call numbers on a single bib is not causing any bloat on my Ubuntu 20.04 test VM. |
| 13:50 |
Dyrcona |
jmurray-isl: I use UDP with rsyslog for "performance reasons." I'd rather lose a few messages on the central logger than have it bog things down. The missing messages are still logged in /var/log/syslog on the host. That said, I've not configured nginx with syslog because I've not felt the need to look at them. I've made do with the Apache logs so far. |
| 13:52 |
jmurray-isl |
Dyrcona: Good to know. I've inherited a system that has it over TCP, and it might be time to change that. Thus far, I'm taking things one step at a time as I can to learn Evergreen. |
| 13:53 |
Dyrcona |
Sure. I think ours was setup with UDP when I got here, but I might have deliberately changed it. That was about 6 years ago, and memory doesn't appear to be serving me right now. :) |
| 14:17 |
Dyrcona |
gmcharlt++ |
| 14:41 |
|
pete joined #evergreen |
| 14:54 |
pinesol |
News from commits: Docs: 3.7.3 Release Notes <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=aa2002eb5b8c3114c3964bd8f73817758a8cf315> |
| 14:55 |
gmcharlt |
Dyrcona++ # thanks for testing! |
| 15:24 |
pinesol |
News from commits: Docs: 3.8.1 Release Notes <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=524164a1a5398a0aa472a5802da175ffe70e63a6> |
| 15:26 |
gmcharlt |
Dyrcona: does CW/MARS use circ limit groups? |
| 15:33 |
Dyrcona |
gmcharlt: I don't know without looking. I'm not really involved with circ configuration. |
| 15:35 |
Dyrcona |
gmcharlt: Apparently, no. I didn't find anything in config.circ_limit_group. |
| 15:35 |
gmcharlt |
thanks for checking |
| 15:36 |
mmorgan |
gmcharlt: NOBLE has a few set up, but I'm really not sure how they're supposed to work. |
| 15:36 |
* mmorgan |
stopped poking when the library was satisfied :) |
| 15:41 |
jihpringle |
mmorgan: same, I can't figure out how they're supposed to work and circ limit sets do what our libraries need |
| 15:43 |
jihpringle |
I was looking at Jeff's fix for circ limit sets and thought it would be good to know what they're supposed to do :) |
| 15:43 |
jihpringle |
https://bugs.launchpad.net/evergreen/+bug/1945385 |
| 15:43 |
pinesol |
Launchpad bug 1945385 in Evergreen 3.8 "Circulation Limit Set interface missing sorting, filters, and back button" [Medium,Confirmed] |
| 15:43 |
jihpringle |
(to know what the groups are supposed to do, not the sets) |
| 15:56 |
gmcharlt |
jihpringle: mmorgan: just responded on the email thread; I think I'm beginning to get a handle on the intention behind the feature, though I'm really curious about what policy scenarios tsbere specifically had in mind when he wrote the feature in 2011 |
| 16:04 |
Dyrcona |
Yeah, I don't have any recollection of that, either. |
| 16:05 |
Dyrcona |
Oh, wait a minute! Reading some of the emails, I think I recall. It had to do with grouping like circulation matrix entries for limits. Like limiting to 5 DVDs including Blu-Ray in the count. |
| 16:07 |
Dyrcona |
Or, maybe, I'm confusing it with limit sets, though the two are probably related. tsbere had a way of doing things in a complicated way to make it universally useful that made sense at the time, but years later seems like overkill. |
| 16:09 |
gmcharlt |
yeah, they're related, but I think the key thing for limit groups is dealing with the cases where circ modifiers and/or shelving locations aren't at the right level of granularity for the limit you want to set |
| 16:09 |
Dyrcona |
Yeah, right. groups could be made and linked to sets to limit items out, etc....because flexibility. |
| 16:10 |
gmcharlt |
yeah |
| 16:10 |
Dyrcona |
Right. |
| 16:10 |
gmcharlt |
and I just came up with this as a possible contender for a succinct description of the feature: ""Circ Limit Groups let you set up policies that assign a tag to loans, then set a limit on the number of active loans with that tag that the patron can have checked out."" |
| 16:10 |
Dyrcona |
I vaguely setting some up all those years ago at MVLC.... I used to know how this worked.... |
| 16:11 |
Dyrcona |
gmcharlt++ : That sounds right to me. |
| 16:11 |
Dyrcona |
I think I omitted "recall" in my previous statement. |
| 16:16 |
* mmorgan |
is trying to remember working on the limits. |
| 16:17 |
mmorgan |
In the circ policies, you can only configure linked limit sets. |
| 16:18 |
gmcharlt |
yeah, the association to a limit group is done in the limit set |
| 16:19 |
gmcharlt |
not directly from policy to limit group |
| 16:20 |
* mmorgan |
probably doesn't have a chance to understand this so late on a Friday afternoon :-/ |
| 16:25 |
pinesol |
News from commits: Update 3.8 Point Release Notes <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=5ac1b03e1cf19e6146df54d46c514d9adec6a3e3> |
| 16:25 |
pinesol |
News from commits: Update 3.7 Point Release Notes <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=a6140937e76c34ab91d165928af88f431ec62ba3> |
| 16:27 |
mmorgan |
Light bulb just came on - I think. The limit groups let you define a kind of meta limit, across more than one limit set. |
| 16:27 |
gmcharlt |
indeed |
| 16:32 |
mmorgan |
We have thus far limited only based on circ modifiers and shelving locations, and the limit sets have accommodated that just fine. |
| 16:45 |
jihpringle |
the piece I still don't get (and maybe I'm missing something) is how you tell Evergreen what the group means. I can link the group to a limit set and from their to a circulation policy but I can't see anyway to tell Evergreen this limit group limits on these parameters |
| 16:45 |
gmcharlt |
there's a bit of bootstrapping |
| 16:45 |
gmcharlt |
the limit group gets named whatever you want it to be |
| 16:46 |
gmcharlt |
but the meaning stems from the circ policies that use a limit set that reference that group |
| 16:47 |
gmcharlt |
so, if some of your policies are based on MARC bib record attributes where circ mods or shelving locations don't help with counting loans for the limit, if the most specific policy is matching on a MARC value, the meaning of the limit group that gets applied is a proxy for that MARC value |
| 16:48 |
gmcharlt |
(or whatever else is the most signficant parts of the circ rule) |
| 16:48 |
jihpringle |
oh, so it's a way to set up a limit that you can apply to a circulation policy that just applies to any items that the policy applies to |
| 16:49 |
gmcharlt |
(which would allow you to do things like set up rules that say somebody who is between the ages of 45 and 56 can check out only 17 items at a time) |
| 16:49 |
jihpringle |
or based on item age |
| 16:49 |
gmcharlt |
yeah |
| 16:49 |
|
jvwoolf left #evergreen |
| 16:50 |
jihpringle |
that makes sense, thanks! |
| 16:50 |
jihpringle |
I'll sum this all up and get it put into the docs :) |
| 16:50 |
mmorgan |
jihpringle++ |
| 16:50 |
gmcharlt |
jihpringle++ |
| 16:50 |
mmorgan |
gmcharlt++ |
| 16:51 |
jihpringle |
gmcharlt++ |
| 17:01 |
|
mmorgan left #evergreen |
| 17:55 |
pinesol |
News from commits: Forward Port 3.7.3 Upgrade Script <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=03485123f1bc148e9371de9e362dec0c7baf1bfe> |
| 18:01 |
pinesol |
News from qatests: Failed Running perl live tests <http://testing.evergreen-ils.org/~live//archive/2022-05/2022-05-20_16:00:02/test.49.html> |
| 18:42 |
|
jihpringle joined #evergreen |
| 20:25 |
pinesol |
News from commits: Forward Port 3.8.1 Upgrade Script <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=15007ec466522c12c51c7ce1844a07a0dd8748bb> |
