| Time |
Nick |
Message |
| 00:21 |
|
jyorio joined #evergreen |
| 01:47 |
|
stryx`7 joined #evergreen |
| 04:20 |
|
joselo joined #evergreen |
| 04:39 |
|
ahills24 joined #evergreen |
| 05:59 |
|
EricZou25 joined #evergreen |
| 07:21 |
|
Snert_15 joined #evergreen |
| 08:19 |
|
mux18 joined #evergreen |
| 08:33 |
|
collum joined #evergreen |
| 08:34 |
|
rlefaive joined #evergreen |
| 08:35 |
|
Dyrcona joined #evergreen |
| 08:38 |
|
kmlussier joined #evergreen |
| 08:38 |
|
mmorgan joined #evergreen |
| 08:53 |
|
lsach joined #evergreen |
| 08:53 |
|
danzilio13 joined #evergreen |
| 09:02 |
|
idjit joined #evergreen |
| 09:02 |
|
jvwoolf joined #evergreen |
| 09:09 |
|
aurieeeh10 joined #evergreen |
| 09:16 |
Dyrcona |
Just checking: A reactor of NOOP_True should do nothing if an action_trigger.event_definition is active, correct? |
| 09:18 |
Dyrcona |
I see "reaction always passes." So, I wonder why a patron got an email from this event? |
| 09:21 |
Bmagic |
ATED was active and reactor was NOOP_True? But it sent an email? What what template? |
| 09:26 |
Dyrcona |
event_defintion.id = 53 |
| 09:27 |
Bmagic |
I mean, was the template column null? |
| 09:30 |
Dyrcona |
The template column is null, but the message_template is not null and that is what the patron received. |
| 09:31 |
|
yboston joined #evergreen |
| 09:33 |
Dyrcona |
I'm starting to think this wasn't an email, but something in the patron message center/whatever-it-is-officially-called. |
| 09:37 |
Dyrcona |
At this point, I'm 96.4% confident that this was not an email. :) |
| 09:40 |
Dyrcona |
And, while on the topic of action_trigger: Does anyone have a strategy for dealing with 2-Day Courtesy notices and the occasional high volume of them? We had 13,868 of them between 4:00 and 5:00 am this morning, and some people get nervous when Nagios says that the number of "events pending" is critical. |
| 09:41 |
Dyrcona |
We probably touched on this around labor day since we had real issues with them that weekend. |
| 09:43 |
Dyrcona |
Wow! Last night was nothing compared to Sunday, the 2nd of September.... http://irc.evergreen-ils.org/evergreen/2018-09-04#i_375131 |
| 09:52 |
csharp |
Dyrcona: what is it about the high volume that's concerning? (we've had resource issues on our server when trying to process many at once, but those have mainly been addressed) |
| 09:53 |
Dyrcona |
csharp: It doesn't bother me, except when it blows up. |
| 09:53 |
csharp |
Dyrcona: by "blows up" do you mean that the A/T process/drone chokes/dies? |
| 09:54 |
Dyrcona |
What else would I mean? :) |
| 09:54 |
Dyrcona |
Actually, I think it may have still been processing events last time I had a real problem. Next time, I will do more investigation before summarily killing processes. |
| 09:54 |
csharp |
we upped our server's RAM and number of processors (on a VM, so that's easy to do) |
| 09:55 |
csharp |
we also tuned opensrf.xml so the trigger service could create enough children |
| 09:55 |
csharp |
and correspondingly cstore procs too (IIRC) |
| 09:55 |
Dyrcona |
My action_triggers run on hardware with 32GB as mentioned in the logs from 2018-09-04. |
| 09:56 |
csharp |
gotcha |
| 09:56 |
csharp |
32GB RAM matches what we have |
| 09:56 |
* csharp |
remembers comparing notes recently |
| 09:56 |
Dyrcona |
Last night wasn't a real problem just the appearance of one given the silly stuff we have in nagios. :) |
| 09:57 |
csharp |
Dyrcona: how many concurrent procs for collect/react do you have set in opensrf.xml? |
| 09:59 |
Dyrcona |
None, apparently. :) |
| 09:59 |
|
jvwoolf left #evergreen |
| 09:59 |
csharp |
we have 3 set for each and that works well for us |
| 09:59 |
Dyrcona |
It's commented out, but the setting is 3. |
| 09:59 |
csharp |
any fewer and we have dying drones |
| 09:59 |
Dyrcona |
So, I assume we're doing 1. |
| 10:00 |
csharp |
we haven't tried higher than 3 |
| 10:00 |
Dyrcona |
I'll try that setting. See if it is an improvement. |
| 10:01 |
Dyrcona |
csharp++ |
| 10:06 |
|
jvwoolf joined #evergreen |
| 10:08 |
Dyrcona |
csharp: Out of curiosity, what are your max_children settings for open-ils.trigger and open-ils.cstore on the vm that runs your triggers? |
| 10:09 |
|
rlefaive joined #evergreen |
| 10:24 |
csharp |
Dyrcona: open-ils.trigger: <max_children>75</max_children> |
| 10:24 |
csharp |
open-ils.cstore: <max_children>128</max_children> |
| 10:24 |
Dyrcona |
csharp: Thanks! That seems like a lot, but I should probably adjust hours upward while I'm at it. |
| 10:25 |
Dyrcona |
heh... "hours..." |
| 10:25 |
csharp |
the cstore setting may not need to be that high - we set it that high to rule it out as a constraint |
| 10:25 |
Dyrcona |
@blame Fingers |
| 10:25 |
pinesol |
Dyrcona: Fingers is the SPY! |
| 10:25 |
csharp |
heh |
| 10:25 |
csharp |
and after it was working, I didn't want to risk problems by setting it lower again :-) |
| 10:26 |
berick |
pg max connections may need upping too if raising cstore a huge amount |
| 10:26 |
csharp |
a/t is awesome when it works and hellish when it doesn't ;-) |
| 10:26 |
csharp |
berick: yeah |
| 10:26 |
Dyrcona |
I dropped the lower settings for cstore from our utility servers, so we're running the same as a single brick drone, but considering that we run two drone servers per brick, I should probably go higher. |
| 10:26 |
csharp |
we've got max connections at 1000 and have never hit that limit that I know of during normal activity |
| 10:26 |
Dyrcona |
berick: While that's true, I've never run into a max connections situation so far. |
| 10:27 |
* Dyrcona |
checks what ours is set to, but I think its about 1,000. |
| 10:27 |
Dyrcona |
Double that: 2,000. |
| 10:27 |
* csharp |
will look into something like pgbouncer if that becomes an issue |
| 10:27 |
csharp |
(based on a suggestion from dbs years ago) |
| 10:28 |
Dyrcona |
Well, our db server is insanely over configured hardware-wise. |
| 10:28 |
Dyrcona |
Well, maybe not "insanely" but definitely over-configured on purpose. |
| 10:28 |
csharp |
after years of underpowered DB hardware, I would always go for overpowered :-) |
| 10:29 |
Dyrcona |
Yeah, 512GB of RAM and SSDs just don't cut it any more....You need 768GB minimum and NVMe drives. :) |
| 10:30 |
Dyrcona |
BTW, ZFS absolutely rocks with PostgreSQL. |
| 10:30 |
Dyrcona |
If I was feeling really bold, I'd switch the O/S on the db server to FreeBSD to get that extra little bit of performance. :P |
| 10:33 |
berick |
do it |
| 10:35 |
* Dyrcona |
imagines berick in a pec flex saying "Do it!" a la Shia Laboeuf. :) |
| 10:36 |
csharp |
we're running it on XFS |
| 10:42 |
Dyrcona |
XFS is cool. I used it on a test db server. |
| 10:43 |
Dyrcona |
The setup time for ZFS is amazing, though. Just run two commands and you're done. No waiting around for a format, etc. |
| 10:44 |
Dyrcona |
Well, you can do it in one, but I wanted a mirrored pool. :) |
| 10:54 |
|
khuckins_ joined #evergreen |
| 11:10 |
|
rlefaive joined #evergreen |
| 11:20 |
|
Christineb joined #evergreen |
| 11:29 |
|
rlefaive joined #evergreen |
| 11:45 |
|
khuckins_ joined #evergreen |
| 12:04 |
|
jihpringle joined #evergreen |
| 12:05 |
|
yboston joined #evergreen |
| 12:08 |
|
khuckins joined #evergreen |
| 12:17 |
|
khuckins_ joined #evergreen |
| 12:26 |
|
khuckins_ joined #evergreen |
| 12:50 |
|
joepie912 joined #evergreen |
| 12:58 |
|
Zorglub joined #evergreen |
| 13:05 |
jonadab |
Hmm. I see not ~/egbuilddir/Open-ILS/src/eg2 nor /openils/Open-ILS/src/eg2 nor /openils/src/eg2 nor any directory named eg2 in any subtree under either ~/egbuilddir nor under /openils/. https://evergreen-ils.org/documentation/install/README_3_2.html#install_files_for_angular_web_staff_client |
| 13:08 |
Dyrcona |
jonadab: "egbuilddir" is the directory where you extracted and are building Evergreen. It's not meant to be taken literally. |
| 13:08 |
jonadab |
Dyrcona: Right, I know. |
| 13:08 |
jonadab |
(The doc actually says $EVERGREEN_ROOT.) |
| 13:08 |
Dyrcona |
OK. |
| 13:08 |
* Dyrcona |
doesn't read docs. :) |
| 13:08 |
|
collum joined #evergreen |
| 13:09 |
bshum |
jonadab: Are you using master? There isn't a rel_3_2 branch around, is there? |
| 13:09 |
Dyrcona |
The first should be there. Are you actually installing 3.2? |
| 13:09 |
jonadab |
egbuilddir/Open-ILS/web/js/ui/default/staff *did* exist, a couple of steps ago. |
| 13:09 |
bshum |
Cause you skip all of section 4 if you're using a tarball |
| 13:09 |
jonadab |
bshum: master branch yes, but probably a little out of date. Let me check that. |
| 13:09 |
Dyrcona |
There isn't a rel_3_2 branch, though there is one for the beta, as well as the tarball. |
| 13:10 |
bshum |
jonadab: Depending on how out of date it is... I think eg2 is new as of the last week and a half or so |
| 13:10 |
jonadab |
Oh, goodness, I'm more out of date than I realized. |
| 13:10 |
Dyrcona |
Pro Tip: Always git fetch --all, git pull before doing any serious work. :) |
| 13:10 |
bshum |
Yeah, a lot of stuff has gone into master the past two weeks |
| 13:10 |
jonadab |
commit 4cd44bb331eb, way back in May. |
| 13:10 |
jonadab |
Let me rectify that. |
| 13:10 |
pinesol |
jonadab: [evergreen|Bill Erickson] LP#1774427 Parse DoB dates as whole dates - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=4cd44bb> |
| 13:11 |
bshum |
So you're following the right steps, just an outdated master :) |
| 13:11 |
jonadab |
Looks like. |
| 13:11 |
Dyrcona |
This is also where using the local README helps. :) |
| 13:12 |
jonadab |
Dyrcona: Good point. |
| 13:18 |
Dyrcona |
That section of the README could use a little clean up for clarity, though. We should probably rename section 4.4 and maybe add some kind of blurb about AngularJS v. Angular. |
| 13:19 |
* Dyrcona |
assigns that ticket to bshum. :P |
| 14:05 |
|
bdljohn1 joined #evergreen |
| 14:06 |
|
bdljohn joined #evergreen |
| 14:08 |
|
nfBurton joined #evergreen |
| 14:21 |
kmlussier |
Are we doing maintenance releases today? |
| 14:22 |
miker |
berick: thanks for looking at the opensrf bug! |
| 14:23 |
Dyrcona |
miker | berick: I never did get either branch/patch to work. |
| 14:24 |
Dyrcona |
kmlussier: I don't know. I asked in the release channel and all I was told was that the 3.2-RC is moved to Monday. |
| 14:25 |
kmlussier |
Dyrcona: Thanks! |
| 14:25 |
kmlussier |
If the date for maintenance releases was also changed, I can update the calendar. |
| 14:44 |
dbwells |
kmlussier: It wasn't considered directly with the RC change, but I vote we push the maintenance releases to next week as well. That way, coming fixes in the RC can potentially make it into 3.1. I'll plan on next Wednesday. Thanks! |
| 14:44 |
kmlussier |
dbwells: Thanks! |
| 14:45 |
Dyrcona |
dbwells: Works for me. |
| 14:45 |
berick |
miker: yep, just a few minor issues so far. making progress. |
| 14:46 |
Dyrcona |
And, someone always beats me to updating the calendar. |
| 14:52 |
|
yboston joined #evergreen |
| 15:03 |
kmlussier |
Dyrcona: That was me. Sorry! |
| 15:03 |
Dyrcona |
I thought it might have been you. |
| 15:21 |
|
khuckins joined #evergreen |
| 15:39 |
|
neutraltone_ joined #evergreen |
| 15:42 |
|
khuckins joined #evergreen |
| 15:54 |
|
bdljohn1 joined #evergreen |
| 15:56 |
|
bdljohn joined #evergreen |
| 15:57 |
|
bdljohn1 joined #evergreen |
| 16:08 |
|
rlefaive joined #evergreen |
| 16:09 |
|
yboston joined #evergreen |
| 16:22 |
Dyrcona |
berick++ |
| 16:37 |
|
rlefaive joined #evergreen |
| 17:02 |
|
dunz0r15 joined #evergreen |
| 17:08 |
|
mmorgan left #evergreen |
| 17:08 |
|
arbrandes26 joined #evergreen |
| 17:17 |
|
Guest84464 joined #evergreen |
| 17:55 |
|
jyap8 joined #evergreen |
| 17:56 |
|
rlefaive joined #evergreen |
| 19:17 |
|
wk-work joined #evergreen |
| 20:09 |
|
skooch joined #evergreen |
| 20:24 |
|
wadeck5 joined #evergreen |
| 20:35 |
|
charz22 joined #evergreen |
| 20:50 |
|
Majadon9 joined #evergreen |
| 21:01 |
|
serflog joined #evergreen |
| 21:01 |
|
Topic for #evergreen is now Welcome to the #evergreen library system channel! | We are publicly logged: http://irc.evergreen-ils.org/evergreen | Large pastes at http://paste.evergreen-ils.org | Can't speak? Make sure your nickname is registered and that you are identified to freenode services: https://freenode.net/kb/answer/registration |
| 21:01 |
|
ericar joined #evergreen |
| 21:01 |
|
rhamby joined #evergreen |
| 21:01 |
|
abneiman joined #evergreen |
| 21:01 |
|
gsams_ joined #evergreen |
| 22:00 |
bshum |
berick++ # clean Ubuntu 18.04 has working opensrf.math for me now! RE: https://bugs.launchpad.net/opensrf/+bug/1793356 |
| 22:00 |
pinesol |
Launchpad bug 1793356 in OpenSRF "Ejabberd strips custom XML attributes" [Medium,Confirmed] - Assigned to Jason Stephenson (jstephenson) |
| 22:01 |
bshum |
Testing the next rounds with Evergreen will require further tweaking and building new Makefile for it. I'll poke at that further and report back when I can. |
| 23:30 |
|
yar joined #evergreen |
