| Time |
Nick |
Message |
| 00:45 |
|
eady joined #evergreen |
| 01:09 |
|
jyorio joined #evergreen |
| 01:09 |
|
akilsdonk joined #evergreen |
| 01:09 |
|
miker joined #evergreen |
| 01:09 |
|
JBoyer joined #evergreen |
| 01:09 |
|
phasefx_ joined #evergreen |
| 01:09 |
|
drigney joined #evergreen |
| 01:09 |
|
ericar joined #evergreen |
| 01:09 |
|
rhamby joined #evergreen |
| 01:09 |
|
felicia joined #evergreen |
| 01:11 |
|
abneiman joined #evergreen |
| 01:11 |
|
jgoodson joined #evergreen |
| 06:20 |
|
sandbergja joined #evergreen |
| 06:28 |
|
remingtron_ joined #evergreen |
| 07:01 |
|
agoben joined #evergreen |
| 07:14 |
|
rjackson_isl joined #evergreen |
| 08:11 |
|
tlittle joined #evergreen |
| 08:39 |
|
mmorgan joined #evergreen |
| 08:43 |
|
jvwoolf joined #evergreen |
| 08:46 |
|
jvwoolf1 joined #evergreen |
| 09:03 |
|
Dyrcona joined #evergreen |
| 09:18 |
Dyrcona |
Ah ha! aacs: action.all_circulation_slim appears to be my culprit. |
| 09:18 |
Dyrcona |
incomplete-features-- |
| 09:18 |
|
cmalm joined #evergreen |
| 09:19 |
Dyrcona |
Fields were added to the view, but not the IDL. |
| 09:19 |
Dyrcona |
auto_renewal fields, that is. |
| 09:21 |
Dyrcona |
I guess I should add them to the end of the fields list? |
| 09:25 |
Dyrcona |
Eh, still no dice.... |
| 09:25 |
* Dyrcona |
checks the view in the database. Maybe it's missing from the upgrade script. |
| 09:26 |
Dyrcona |
Well, the fields are there. |
| 09:26 |
pinesol |
[evergreen|Remington Steed] Docs: Make "staff catalog" titles consistent in 3.4 Rel Notes - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=fe1c37a> |
| 09:26 |
pinesol |
[evergreen|Remington Steed] Docs: Group similar release notes together - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=0cb51a2> |
| 09:27 |
Dyrcona |
Clear site data and give it another shot.... |
| 09:29 |
Dyrcona |
Nope. still busted. |
| 09:30 |
Dyrcona |
So, must be some other class... |
| 09:31 |
Dyrcona |
And, this illustrates how inefficiently the webstaff client retrieves different versions of the same data over and over again, 'cause it's already grabbed combcirc and then it goes for aacs. I guess I'll find out what else it grabs on the next run through. |
| 09:33 |
Dyrcona |
Before trying with the debugger again, I'm going to do the install dance, autogen.sh, clear site data, and give it one more go. |
| 09:33 |
mmorgan |
Dyrcona: Which grid are you looking at? |
| 09:33 |
Dyrcona |
Item Status Detail View Quick Summary and Circ History List |
| 09:33 |
|
nfBurton joined #evergreen |
| 09:35 |
Dyrcona |
Oh, great. Now I can't login.... |
| 09:36 |
Dyrcona |
heh.. musta typoed the password, twice in a row. :) |
| 09:37 |
Dyrcona |
Yeah, still not working, so must be some other class..... :( |
| 09:49 |
mmorgan |
Dyrcona: does bug 1787415 help? circ_summary uses class accs |
| 09:49 |
pinesol |
Launchpad bug 1787415 in Evergreen 3.3 "Most Recent Circ Group Inconsistent Data Displayed" [Undecided,Confirmed] https://launchpad.net/bugs/1787415 |
| 09:50 |
Dyrcona |
systemd-resolvd-- |
| 09:50 |
Dyrcona |
It might, if Lp would load because of systemd.... :( |
| 09:55 |
Dyrcona |
mmorgan: Thanks, but that's not it. I'm trying to figure out what class "circ" actually is on those templates. It's evidently not circ because that object has auto_renewal working just fine. I thought it was combcirc based on my first debugger run, but no, apparently not. |
| 09:55 |
Dyrcona |
Then, it looked like aacs after my second run, but no, apparently not that one, either. |
| 09:56 |
Dyrcona |
webstaff-client-- |
| 10:04 |
dbs |
Dyrcona: Over the weekend, I participated in a brief collective effort to customize a template for an organizational website that consisted of roughly 5 pages |
| 10:04 |
dbs |
'npm install' downloaded over 1,000 packages |
| 10:05 |
Dyrcona |
So, you're telling me it's a bug in a dependency, or might be? |
| 10:05 |
dbs |
Methinks there was a serious bit of overengineering going on for that template! In comparison, at least the web staff client has a good rationale for its complexity |
| 10:07 |
dbs |
Well I guess it could be a dependency bug. I was just ruing the state of modern web dev, where a mostly static set of pages starts with 1,000 npm packages (500 of which were flagged for security vulnerabilities, as a bonus) |
| 10:07 |
Dyrcona |
Well... That's debatable.... Too much use of PCRUD to my liking. |
| 10:07 |
Dyrcona |
Well, yeah, web development is in a terrible state in general. |
| 10:08 |
Dyrcona |
So, I'm pretty sure that "circ" on the summary pane is aacs. |
| 10:09 |
Dyrcona |
aacs was missing auto_renewal from the IDL, but after fixing that, it still isn't showing up.... |
| 10:09 |
dbs |
I guess pcrud gives us a nice generic approach to do almost everything, with the idea that large/costly/frequent paths would warrant spawning a dedicated API that would speed up & simplify some things? |
| 10:10 |
berick |
dbs: exactly |
| 10:10 |
Dyrcona |
Right, but using pcrud in lieu of back end API is a bad idea. PCRUD is a database access mechanism, not an API. |
| 10:10 |
Dyrcona |
So, berick, why can't I get Renewal Type to say "Automatic?" |
| 10:12 |
Dyrcona |
Am I missing something after updating the IDL, doing npm run build-prod, make install, chown, restart, etc.? |
| 10:12 |
berick |
Dyrcona: where's your branch? |
| 10:12 |
Dyrcona |
Let me add a commit and I'll share the name again. I'll fix the IDL entry for aacs which hasn't made it to my branch, yet. |
| 10:14 |
Dyrcona |
working/user/dyrcona/lp1835085-auto_renewal-not-d |
| 10:14 |
Dyrcona |
Oops. cut off. |
| 10:14 |
Dyrcona |
working/user/dyrcona/lp1835085-auto_renewal-not-desk-renewal |
| 10:15 |
Dyrcona |
I also have this one installed working/user/dyrcona/lp1835577-add-missing-autorenewal-IDL-fields |
| 10:16 |
Dyrcona |
And maybe the top commit should get moved, but whatevs.... |
| 10:18 |
Dyrcona |
copy and pasted with the correct name this time: working/user/dyrcona/lp1835085-auto_renewal-not-desk_renewal |
| 10:18 |
Dyrcona |
typing-- |
| 10:21 |
* Dyrcona |
is frustrated. Don't mind his mild rantings... |
| 10:22 |
Dyrcona |
But, I don't understand why we grab aacs if combcirc comes with the copy? I'm ready to be enlightened. |
| 10:30 |
berick |
Dyrcona: http://i.imgur.com/6m7g9y2.png |
| 10:30 |
berick |
i didn't make any changes |
| 10:31 |
berick |
except to force a circ to be an auto_renewal=true |
| 10:32 |
Dyrcona |
Well, that compounds my frustration, because it means something is hosed on my training server more than I suspected. |
| 10:32 |
Dyrcona |
berick++ for looking |
| 10:32 |
Dyrcona |
dbs++ for commiserating |
| 10:32 |
Dyrcona |
mmorgan++ for suggesting |
| 10:33 |
Dyrcona |
Thing is everything looks good on my end, but it just doesn't work. |
| 10:34 |
Dyrcona |
berick: Should move the IDL commit to my other branch? I think putting the IDL fixes in one place is better. |
| 10:37 |
berick |
i'd probably leave it in both and comment to that affect in LP1835085 |
| 10:39 |
Dyrcona |
OK. |
| 10:41 |
phasefx_ |
random aside, the live tests for today did finish, but the ssh call to invoke installer_installer.sh remotely is hanging |
| 10:41 |
Dyrcona |
I'm going to check a few more things on the server, make sure the IDL is valid, etc.... |
| 10:41 |
|
Dyrcona joined #evergreen |
| 10:42 |
phasefx_ |
all tests passed |
| 10:42 |
Dyrcona |
between resold and my wifi, It's a wonder I can do anything. |
| 10:42 |
Dyrcona |
and typos, yay! :) |
| 10:43 |
Dyrcona |
phasefx: good news. |
| 10:43 |
phasefx |
we'll see if it hangs again this afternoon; I just killed the ssh process and it finished as normal |
| 10:44 |
berick |
Dyrcona: easy quick check is https://hostname/IDL2js |
| 10:44 |
berick |
ctrl-f 'auto_renewal' |
| 10:45 |
berick |
i'm seeing 5 matches |
| 10:45 |
Dyrcona |
berick: Thanks! I always forget that. |
| 10:46 |
Dyrcona |
I find 9, but I have that other branch installed. |
| 10:48 |
Dyrcona |
I think auto_renewal and auto_renewal_remaining may be missing from other classes. If I search "desk_renewal" I find 6. If I search "auto_renewal" I find 4. |
| 10:50 |
|
tlittle32 joined #evergreen |
| 10:52 |
Dyrcona |
These might be custom IDL classes, btw. I haven't checked, but we have a few custom classes for reports. |
| 10:56 |
|
rjackson_isl joined #evergreen |
| 10:58 |
|
tlittle47 joined #evergreen |
| 10:59 |
Dyrcona |
action.open_circulation and circbyyr need auto_renewal added. |
| 11:00 |
* Dyrcona |
will add them to the IDL branch. |
| 11:01 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 11:04 |
|
Christineb joined #evergreen |
| 11:11 |
Dyrcona |
This seems weird to me. The action.open_circulation view is defined as select * from action.circuation where checkin_time is null order by due_date. However, auto_renewal and auto_renewal_remaining don't show up in the column list. Does it need to be recreated, I wonder? |
| 11:13 |
Dyrcona |
Same thing with action.billable_circulations.... |
| 11:13 |
berick |
yeah, the views have to be rebuilt |
| 11:13 |
Dyrcona |
I'll add that to the IDL branch. |
| 11:14 |
Dyrcona |
I just rebuilt action.open_circulation. |
| 11:32 |
* mmorgan |
is looking again at my paste from last week http://paste.evergreen-ils.org/10051 |
| 11:33 |
|
rjackson_isl_ joined #evergreen |
| 11:33 |
|
agoben_ joined #evergreen |
| 11:33 |
mmorgan |
I'm seeing this same sort of thing several times in an hour. |
| 11:36 |
mmorgan |
I'm wondering if these log entries correspond to catalogers' reports of copy templates suddenly disappearing from the display and the screen 'freezing' |
| 11:36 |
mmorgan |
Anyone seen anything similar? |
| 11:39 |
mmorgan |
I notice that there's a call for the user's webstaff.cat.copy.templates, followed by a call for staff_client.copy_editor.templates. |
| 11:40 |
mmorgan |
Should the xul templates get retrieved if web client templates exist? |
| 11:41 |
Dyrcona |
No, but if there's a timeout trying to retrieve the web staff templates it will look like they don't exist. |
| 11:42 |
|
khaun joined #evergreen |
| 11:45 |
Dyrcona |
Has anyone gotten reports of Z39.50 searches, particularly of OCLC, being slow lately? |
| 11:48 |
|
jvwoolf joined #evergreen |
| 11:48 |
csharp |
nothing on our cataloging email list, which is where that kind of complaint usually shows up |
| 11:50 |
|
jihpringle joined #evergreen |
| 11:52 |
|
aabbee joined #evergreen |
| 11:55 |
JBoyer |
mmorgan, it tries to retrieve both just in case. That may have been a shortsighted simplification on my part because if you have huge templates that nearly doubles their size. :( |
| 11:55 |
JBoyer |
(I seriously underestimated some people's template use) |
| 11:57 |
Dyrcona |
csharp: Thanks! I'll assume it's just us for now. |
| 12:00 |
mmorgan |
JBoyer: Ok, thanks, I might try removing the staff_client templates for a particular user to see if it makes a difference. |
| 12:31 |
|
collum joined #evergreen |
| 12:49 |
|
khuckins joined #evergreen |
| 12:49 |
Dyrcona |
berick: After applying all of the additional IDL and view fixes, and doing a full rebuild of Evergreen, I'm getting the auto_renewal fields to work as I intended in the Item Status views. Thanks, again, for your patient help earlier today. |
| 12:50 |
Dyrcona |
berick++ |
| 12:50 |
berick |
Dyrcona: yay, glad you got it working |
| 12:51 |
Dyrcona |
I think that maybe a "make" was required. |
| 12:57 |
|
collum_ joined #evergreen |
| 13:03 |
pinesol |
[evergreen|blake] LP1735847: Remove patron info from default Hold Transit Slip - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=10e7756> |
| 13:19 |
gmcharlt |
the 3.4 beta is now available from https://evergreen-ils.org/egdownloads/ |
| 13:22 |
Dyrcona |
gmcharlt++ |
| 13:25 |
csharp |
gmcharlt++ |
| 13:34 |
|
yboston joined #evergreen |
| 13:43 |
jeff |
surprised it took this long for us to run into bug 1843466 |
| 13:43 |
pinesol |
Launchpad bug 1843466 in Evergreen "SIP crash with metarecord holds and msg64_hold_datatype = barcode" [Undecided,New] https://launchpad.net/bugs/1843466 - Assigned to Jeff Godin (jgodin) |
| 13:44 |
jeff |
but we don't use metarecord holds much (and have them turned off for patrons, mostly) |
| 13:45 |
|
collum joined #evergreen |
| 13:57 |
|
bos20k joined #evergreen |
| 14:26 |
mmorgan |
reproducibility++ |
| 14:27 |
mmorgan |
I was able to reproduce the NOT CONNECTED log entries by copying the xul client and web client copy template user settings to a test user. After clicking Add Holdings on a few bib records, I saw the error :) |
| 14:28 |
mmorgan |
The xul templates are BIG |
| 14:34 |
csharp |
@who is NOT CONNECTED TO THE NETWORK!!!? |
| 14:34 |
pinesol |
gsams__ is NOT CONNECTED TO THE NETWORK. |
| 14:35 |
* mmorgan |
also often feels that way :) |
| 14:35 |
berick |
best way to avoid Cylons |
| 14:35 |
csharp |
@who is a frakkin' toaster? |
| 14:35 |
pinesol |
genpaku is a frakkin' toaster. |
| 14:36 |
csharp |
currently on my *mumbles*th rewatch of that show - one of my favs |
| 14:37 |
berick |
oh that first season is something special |
| 14:37 |
* berick |
just rewatched the episode of Portlandia where they binge Battlestar :) |
| 14:46 |
csharp |
berick: also a favorite at our house :-) |
| 14:47 |
|
khuckins joined #evergreen |
| 14:48 |
Dyrcona |
So, since we set up the nginx proxy on the brickheads things have gotten better and worse. We don't get reports of white screens any more, but catalogers claim that Evergreen is now so slow as to be practically unusable. |
| 14:49 |
Dyrcona |
I also think it worth noting that Syrup stopped working when pointing at the bricks. I had to tell it to use the gateway on one of the utility servers that doesn't run the proxy, and only runs Apache for z39.50. |
| 14:50 |
Dyrcona |
I think the z39.50 issues that I mentioned earlier may be related. |
| 14:50 |
Dyrcona |
ldirectord also seems to crash on a weekly basis after the proxy set up. |
| 14:50 |
agoben_ |
Can someone point me toward whoever has permission to create accounts for the evergreen-ils.org wiki? |
| 14:51 |
Dyrcona |
agoben_: dbs set me up all those years ago. |
| 14:51 |
agoben_ |
Dyrcona++ Thanks! |
| 14:51 |
csharp |
I apparently don't have that power (not that I need it) |
| 14:51 |
csharp |
agoben_: bshum and gmcharlt and maybe phasefx ? |
| 14:52 |
gmcharlt |
more broadly, docs evergreen-ils.org |
| 14:52 |
agoben_ |
Oh, I actually have a couple of those emails a bit closer to hand, thanks! |
| 14:52 |
agoben_ |
csharp++ |
| 14:52 |
agoben_ |
gmcharlt++ |
| 14:54 |
|
khuckins_ joined #evergreen |
| 14:57 |
Dyrcona |
We get all the "fun" problems..... |
| 15:02 |
Dyrcona |
csharp: I'm curious. Could you run ipvsadm -l on your load balancer and add the active and inactive https connections. I'd like to compare numbers with our 2,309. |
| 15:02 |
|
mmorgan1 joined #evergreen |
| 15:30 |
csharp |
Dyrcona: https://pastebin.com/gcqjAAEt |
| 15:30 |
bshum |
That's pretty quiet |
| 15:31 |
bshum |
Comparatively speaking |
| 15:31 |
Dyrcona |
csharp: Thanks. I mistyped earlier. That should have 1,309 https connections. Your numbers look similar to our. |
| 15:32 |
csharp |
those numbers used to be far higher pre-migration |
| 15:33 |
csharp |
speedy servers mean fewer connections in my experience |
| 15:33 |
Dyrcona |
csharp: Could I trouble you to run "ps -o %mem,rsz,sz,vsz $pid" where $pid is the pid of your ldirectord main process? |
| 15:34 |
csharp |
%MEM RSZ SZ VSZ |
| 15:34 |
csharp |
0.1 29440 24327 97308 |
| 15:34 |
Dyrcona |
Do you have two running? Could you run it on both if so? |
| 15:34 |
Dyrcona |
I'll paste our results for the logs. It will flabbergast you. |
| 15:35 |
Dyrcona |
%MEM RSZ SZ VSZ |
| 15:35 |
Dyrcona |
67.8 11051840 2778887 11115548 |
| 15:35 |
csharp |
we have just the one with another in place for failover (connected via heartbeat/corosync) |
| 15:35 |
csharp |
oh wow |
| 15:35 |
Dyrcona |
Oh, it's gone up... |
| 15:35 |
Dyrcona |
Yeah, 10GB roughly speaking. |
| 15:36 |
Dyrcona |
Does pgrep -af ldirectord output 1 or 2 lines? |
| 15:36 |
csharp |
1 line |
| 15:36 |
Dyrcona |
Ok. so does mine. |
| 15:36 |
Dyrcona |
We have 2 if it was started by systemd. |
| 15:37 |
Dyrcona |
So something is seriously wrong here. |
| 15:39 |
|
mmorgan joined #evergreen |
| 16:04 |
|
nfBurton joined #evergreen |
| 17:11 |
|
mmorgan left #evergreen |
| 17:35 |
|
Dyrcona joined #evergreen |
| 18:07 |
|
alynn26 left #evergreen |
| 21:25 |
|
remingtron_ joined #evergreen |
| 22:02 |
|
book` joined #evergreen |
| 22:16 |
|
book`_ joined #evergreen |
| 23:01 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 23:52 |
|
jvwoolf joined #evergreen |
