| Time |
Nick |
Message |
| 06:01 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 07:17 |
|
rjackson_isl_hom joined #evergreen |
| 08:01 |
|
Dyrcona joined #evergreen |
| 08:10 |
|
mantis2 joined #evergreen |
| 08:29 |
|
rfrasur joined #evergreen |
| 08:36 |
|
mmorgan joined #evergreen |
| 08:36 |
Dyrcona |
I'm going through some of my old scripts, mostly things I wrote to fix a specific problem, and I don't remember writing half of them. |
| 08:38 |
mmorgan |
Dyrcona: Must've been your other self :) |
| 08:40 |
Dyrcona |
Indeed! |
| 08:40 |
Dyrcona |
But, I will excuse myself for not remembering some specific thing from three or more years ago. |
| 08:41 |
mmorgan |
:) |
| 08:41 |
Dyrcona |
I suppose I could delete some of these. I guess I kept them thinking that I might either need them again or want to look at them for some reason. |
| 08:45 |
Dyrcona |
Some of them look like I was just playing around with command or functions. |
| 08:49 |
Dyrcona |
Yeah, experiments, tests, one shots, and things that I used to years ago, but are no longer needed. |
| 08:58 |
Dyrcona |
I could organize this particular set of files a bit better. I'm finding some things that will be useful for our next upgrade, I guess because I wrote them for a previous upgrade. |
| 09:07 |
csharp |
Dyrcona: yeah, I have a lot of those fix-and-forget scripts that I've kept around too |
| 09:07 |
csharp |
I try to name them something that I might have a chance of remembering later |
| 09:18 |
Dyrcona |
Yeah, same here, but I often don't remember, and when there all in a directory with a hundred other files..... |
| 09:18 |
Dyrcona |
I found a useful script for deleting and updating node, though. That will come in handy when we upgrade to 3.6. |
| 09:19 |
Dyrcona |
I've files going back to 2008, though some could be older depending on how they were moved/copied around. |
| 09:40 |
|
rfrasur joined #evergreen |
| 10:14 |
|
rlefaive joined #evergreen |
| 10:24 |
csharp |
release_team++ |
| 10:25 |
|
sandbergja joined #evergreen |
| 10:25 |
JBoyer |
Dyrcona++ |
| 10:25 |
JBoyer |
csharp++ |
| 10:25 |
Dyrcona |
csharp++ JBoyer++ sandbergja++ |
| 10:26 |
JBoyer |
sandbergja++ |
| 10:26 |
sandbergja |
yay releases |
| 10:33 |
|
Cocopuff2018 joined #evergreen |
| 10:40 |
Dyrcona |
all to [open-ils.circ.renew] failed for session [1606316647.9015115338.4185859235], thread trace [1]:#012Can't call method "circ_lib" on an undefined value at /usr/local/share/perl/5.26.1/OpenILS/Application/Circ/Circulate.pm line 1035. |
| 10:41 |
Dyrcona |
On training, and it makes no sense because that line is $self->circ_lib and just two lines above that another method is called on $self, and $self doesn't get reassigned in the meantime. |
| 10:43 |
Dyrcona |
find /usr/local -name Circulate.pm |
| 10:43 |
Dyrcona |
/usr/local/share/perl/5.26.1/OpenILS/Application/Circ/Circulate.pm |
| 10:43 |
Dyrcona |
/usr/local/share/perl/5.22.1/OpenILS/Application/Circ/Circulate.pm |
| 10:44 |
|
rfrasur joined #evergreen |
| 10:45 |
Dyrcona |
And, that's not my issue. It's $self->circ->circ_lib later on the same line. |
| 10:49 |
|
sandbergja joined #evergreen |
| 10:51 |
pinesol |
[evergreen|Jason Boyer] Correct asciidoc issues - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=c491e97> |
| 12:05 |
|
jihpringle joined #evergreen |
| 13:01 |
|
Christineb joined #evergreen |
| 14:00 |
|
sandbergja joined #evergreen |
| 14:10 |
Bmagic |
has anyone seen this error before when running autogen : [ERR :4594:Server.pm:585:] server: child process died: XPath error : Invalid expression /opensrf/default|/opensrf/hosts/{my-server-hostname} |
| 14:16 |
Bmagic |
I left the first par to of that off on accident "opensrf.settings" |
| 14:27 |
Dyrcona |
Bmagic: No, but I'd check the format of opensrf_core.xml and opensrf.xml, first. |
| 14:28 |
Bmagic |
good idea. I was reading Server.pm from the OpenSRF repo |
| 14:28 |
jeff |
literally {my-server-hostname}? |
| 14:28 |
jeff |
or was the error along the lines of XPath error : Invalid expression/opensrf/default|/opensrf/hosts/evergreen.example.com |
| 14:29 |
Bmagic |
jeff: no, that part of the log message I sanitized |
| 14:30 |
Bmagic |
right, evergreen.example.com |
| 14:30 |
Dyrcona |
Is it the |? |
| 14:30 |
Bmagic |
Dyrcona: those configs are identical to production bricks |
| 14:31 |
Bmagic |
The rest of the error message is untouched |
| 14:31 |
Dyrcona |
Eh, no. The pipe means or. |
| 14:32 |
jeff |
new failure, or first attempt on a new host? |
| 14:32 |
Bmagic |
something is tricking OpenSRF. It is a container and an 18.04 container |
| 14:32 |
jeff |
has the relevant XPath lib changed its tolerance for something that used to work? |
| 14:32 |
Bmagic |
new machine with new host OS (CentOS 7+ host, 18.04 container) |
| 14:32 |
jeff |
nothing weird in your hostname, is there? |
| 14:33 |
Bmagic |
hostname is normal looking. though it does start with a number if that matters? |
| 14:33 |
Dyrcona |
Does the hostname contain an @ sign? |
| 14:33 |
Bmagic |
no |
| 14:34 |
Dyrcona |
I've heard of people having issues with other software using docker, sometimes weird hostnames. |
| 14:34 |
pastebot |
"Bmagic" at 168.25.130.30 pasted "Full error" (15 lines) at http://paste.evergreen-ils.org/10058 |
| 14:34 |
Bmagic |
That's the complete lead up to the error |
| 14:35 |
Dyrcona |
Well, the caret suggests XPATH doesn't like the hostname. |
| 14:36 |
Bmagic |
right.... but really? |
| 14:36 |
Dyrcona |
I don't know why, unless there's a rule about xml tags not starting with numbers. |
| 14:37 |
Dyrcona |
Bingo! |
| 14:37 |
Dyrcona |
Element names must start with a letter or underscore. |
| 14:37 |
Bmagic |
dang.... I could have sworn I have containers all over the place that start with numbers |
| 14:38 |
Bmagic |
ok, changing host name and see if I can get passed that error |
| 14:42 |
Bmagic |
Dyrcona++ jeff++ |
| 14:42 |
Bmagic |
thanks for listening... it's working with a different hostname. Wild. Been using containers for 4 years.... hard to believe it's never randomized a hostname starting with a number before.. |
| 14:47 |
Dyrcona |
Is the first one you've installed with Ubuntu 18.04? If so, it could be that previous versions of the XPath software were not enforcing that rule. |
| 15:34 |
Bmagic |
that sounds more like the cause |
| 15:40 |
|
mantis2 left #evergreen |
| 16:16 |
mmorgan |
I have a test server running (finally!) and am testing making changes to Angular files. I'm not seeing my changes in the web client. |
| 16:17 |
mmorgan |
I make a change, do the ng build --prod to compile, but I'm missing the step(s) to get the compiled files where they belong. |
| 16:17 |
berick |
mmorgan: copy them to /openils/var/web/eg2/en-US/ |
| 16:18 |
berick |
and copy them from your checkout at Open-ILS/web/eg2/en-US/ |
| 16:18 |
mmorgan |
berick: Just copy them? Thought I had tried that, but will try again. |
| 16:19 |
berick |
yeah, should be fine to just copy them over |
| 16:21 |
mmorgan |
berick: Do services need restarting after the copy? |
| 16:21 |
berick |
mmorgan: nope, but a browser cache clear or hard refresh likely will be |
| 16:22 |
mmorgan |
Ok, good to know. Giving it a try. |
| 16:29 |
mmorgan |
Hmm. Not seeing the changes. Using bug 1855761 to test. Still seeing the typo 'Succeessfully' |
| 16:29 |
pinesol |
Launchpad bug 1855761 in Evergreen "Typo when creating new carousel mapping" [Low,Confirmed] https://launchpad.net/bugs/1855761 |
| 16:29 |
|
sandbergja joined #evergreen |
| 16:44 |
mmorgan |
Oh. 'Succeessfully' appears in other files, too. I do see a result from my changes, though, so |
| 16:44 |
mmorgan |
berick++ |
| 16:44 |
mmorgan |
Thanks |
| 16:44 |
mmorgan |
! |
| 16:47 |
|
rlefaive joined #evergreen |
| 16:58 |
berick |
mmorgan: yay |
| 17:04 |
csharp |
@decide success or succeess |
| 17:04 |
pinesol |
csharp: go with success |
| 17:04 |
mmorgan |
:) |
| 17:04 |
mmorgan |
Though, I'll take either! Both are better than faiilure! |
| 17:20 |
jeffdavis |
"no children available" for open-ils.actor even after raising max children to 200 |
| 17:23 |
jeffdavis |
including 112 calls to open-ils.actor.settings.apply.user_or_ws for the same authtoken over a single 6-second period |
| 17:26 |
jeffdavis |
The user appears to have simply been checking in items. |
| 17:42 |
jeffdavis |
Initial hypothesis: circ/checkin/app.js > getItemBatch > getServerItemBatch > handleServerItemResponse > setServerItem |
| 17:42 |
jeffdavis |
hatch.setServerItem is the only function that calls open-ils.actor.settings.apply.user_or_ws in AngJS |
| 17:43 |
jeffdavis |
looks to me like hatch.getServerItemBatch will iterate through a list of settings and call handleServerItemResponse for each one, which calls setServerItem (and thus does the user_or_ws open-ils.actor call) for each setting |
| 17:43 |
jeffdavis |
(potentially) |
| 17:44 |
jeffdavis |
(if there are local settings to be migrated to the server?) |
| 17:50 |
jeffdavis |
hm, my "no children available" errors precede the user_or_ws calls by a couple seconds, maybe all those calls are because the service was unavailable? i.e. client tries to lookup server-side settings via getItemBatch, this fails because no open-ils.actor drones are available, so assumes the settings don't exist server-side and therefore tries to migrate them with the user_or_ws calls |
| 17:50 |
jeffdavis |
That would explain why this starts showing up in the middle of an ordinary checkin session, but doesn't explain how we got to "no children available" in the first place. |
| 17:53 |
|
mantis2 joined #evergreen |
| 17:53 |
|
mantis2 left #evergreen |
| 17:56 |
|
mmorgan left #evergreen |
| 18:01 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 18:06 |
jeffdavis |
200 max children for open-ils.actor, "no children available" starting at 14:03:15.146000, no sudden spike in open-ils.actor calls beforehand and only 269 such calls in the 10 mins leading up to it |
| 18:08 |
jeffdavis |
lots of open-ils.actor NOT CONNECTED TO THE NETWORK errors immediately afterward, starting at 14:03:15.435596 (almost exactly 0.3s after the first "no children available" error), which seems like a consequence of no children available |
| 18:12 |
jeffdavis |
actor drone count was 21/200 at 14:03:04 and 200/200 at 14:03:19 |
| 18:17 |
jeffdavis |
some of that could be the 112 user_or_ws calls from a single user, but those don't begin until 14:03:17.078026 according to the logs |
| 18:29 |
jeffdavis |
ooh, lots of this kind of thing in gateway logs: [ERR :24464:osrf-websocket-stdio.c:671:] WS received XMPP error message in response to thread=[...] and recipient=router public.localhost/open-ils.pcrud. Likely the recipient is not accessible/available. |
| 18:35 |
jeffdavis |
starting at Likely the recipient is not accessible/available |
| 18:35 |
jeffdavis |
*starting at 14:01:56 |
| 18:36 |
jeffdavis |
keep spammin' those irc logs jeffdavis! |
| 19:46 |
|
rlefaive joined #evergreen |
