| Time |
Nick |
Message |
| 01:49 |
|
sandbergja joined #evergreen |
| 06:01 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 06:57 |
|
jeff joined #evergreen |
| 06:59 |
|
berick joined #evergreen |
| 06:59 |
|
cesardv joined #evergreen |
| 07:29 |
|
rjackson_isl_hom joined #evergreen |
| 07:46 |
|
mantis1 joined #evergreen |
| 08:12 |
csharp |
hmm - this looks wrong: 2021-01-19 08:01:51 db01 postgres[36855]: [371-2] 2021-01-19 08:01:51.888 EST [36855] evergreen evergreen STATEMENT: -- bib search: #CD_documentLength #CD_meanHarmonic #CD_uniqueWords core_limit(10000) badge_orgs(1) estimation_strategy(inclusion) skip_check(0) check_limit(1000) |
| 08:12 |
csharp |
icon_format(book;metarecord=2237848;modifier=metabib;fi:from_metarecord=2237848;expand=marchtml;expand=summaryplus;qtype=subject;query=Mutism Fiction.;pane=advanced;no_highlight=1;qtype=subject;query=Surrogate motherhood Fiction.) depth(0) |
| 08:14 |
csharp |
a different one, full query: https://pastebin.com/Z1L2FgnK |
| 08:14 |
csharp |
I've seen this before, but seems more common today after upgrading to 3.6/bootstrap opac |
| 08:15 |
csharp |
something somewhere needs to do a better job of sanitizing input? |
| 08:29 |
|
mmorgan joined #evergreen |
| 08:45 |
JBoyer |
That looks vaguely familiar. I think it's because there was a time when we tried to use semicolons to split query params rather than ampersands and something still sends that to us. |
| 08:45 |
JBoyer |
The sanitizing should also be smarter but I think there's a fair chance some of this is our own doing. |
| 08:46 |
JBoyer |
I've only ever noticed it in relation to metabib values, but haven't had time to track it down. |
| 08:55 |
|
rjackson_isl_hom joined #evergreen |
| 08:55 |
|
Dyrcona joined #evergreen |
| 08:56 |
|
collum joined #evergreen |
| 09:16 |
csharp |
JBoyer: thanks - if things don't blow up too badly with libraries opening this morning, I'll see what I can figure out |
| 09:16 |
csharp |
agreed that it only appears to be metabib searches |
| 09:29 |
Dyrcona |
We have email checkout receipt events that are going to complete but are not sending emails, and I have not been able to figure it out so far. Has anyone else seen this? |
| 09:31 |
Dyrcona |
Other events using the SendEmail reactor seem to be working OK. |
| 09:46 |
Dyrcona |
I'm not finding any error messages anywhere, either. |
| 10:01 |
Dyrcona |
The template output looks OK when compared to other events that use the SendEmail reactor. |
| 10:03 |
Dyrcona |
Oh. I think I get it. The emails are not sent from the machine that does the a/t jobs. It's sent from the bricks...... |
| 10:04 |
Dyrcona |
So, my bricks all have to be able to send email..... |
| 10:06 |
Dyrcona |
At least, a) that's what it looks like and b) this has a familiar smell. |
| 10:11 |
Dyrcona |
Also, don't remember if it's the brick head or the drones that need it. |
| 10:11 |
Dyrcona |
Guess, I'll configure them all for smarthost through the utility server. |
| 10:19 |
mmorgan |
leafblowers-- |
| 10:19 |
mmorgan |
Especially in the middle of January |
| 10:24 |
csharp |
https://twitter.com/MrDys/status/1329085491786911746 |
| 10:26 |
Dyrcona |
Heh. And who says Twitter is a complete waste of time? ... Oh right, I do. :) |
| 10:26 |
mmorgan |
csharp++ |
| 10:27 |
Dyrcona |
I gave csharp karma twice yesterday, so it will have to wait. :) |
| 10:28 |
csharp |
please don't go over my quota |
| 10:28 |
Dyrcona |
@quota csharp |
| 10:28 |
pinesol |
Dyrcona: Down time is a fact of business when you're a poor 501c3 corporation. |
| 10:28 |
Dyrcona |
csharp++ |
| 10:53 |
Dyrcona |
@quote random |
| 10:53 |
pinesol |
Dyrcona: Quote #199: "<berick> love there's an EDI code for number of law suits" (added by Dyrcona at 06:18 PM, May 28, 2019) |
| 11:00 |
csharp |
something in staff catalog search is generating NOT CONNECTED errors |
| 11:01 |
csharp |
https://pastebin.com/yuU0ZQVB - full logs for one of them |
| 11:07 |
Dyrcona |
I gave up on 'not connected' errors some time ago. |
| 11:07 |
Dyrcona |
Search seems to be a huge source, I suspect timeouts. |
| 11:08 |
Dyrcona |
Going back to my "monologue" about email checkout receipts, it turns out that the drone servers, if you have them, need to be able to send email, preferably to a smart relay. |
| 11:10 |
|
Cocopuff2018 joined #evergreen |
| 11:37 |
|
jihpringle joined #evergreen |
| 11:51 |
Dyrcona |
The staler your data, the longer the fine generator runs. |
| 11:51 |
|
cesardv joined #evergreen |
| 11:59 |
|
sandbergja joined #evergreen |
| 12:04 |
csharp |
whatever those NOT CONNECTED errors are (not) doing doesn't seem to be generating helpdesk tickets - not sure if that it's something that works the second time or what |
| 12:08 |
JBoyer |
csharp, it's possible those are apache clients just "hanging up" if the end user browses away or otherwise doesn't wait for the search to finish. |
| 12:09 |
JBoyer |
You might be able to try that out by searching for something unusual that doesn't immediately return 0 results, let it spin a couple seconds, and then just close the tab. If that's the situation you should be able to follow along in the logs. |
| 12:14 |
Dyrcona |
"Why can't it work like Google?" "I don't know. Give me a hundred billion dollars, and I'll get back to you." |
| 12:15 |
Dyrcona |
"A hundred million ought to be enough to fix anything." <-- Guess who. |
| 12:58 |
abneiman |
Dyrcona++ |
| 13:01 |
|
jvwoolf joined #evergreen |
| 13:04 |
|
jvwoolf1 joined #evergreen |
| 13:10 |
|
sandbergja joined #evergreen |
| 13:19 |
Dyrcona |
Hmm.. Since I'm "downgrading" a test machine from 3.6.1 to 3.5.2, I guess I'll wipe out /openils. |
| 13:31 |
csharp |
JBoyer++ |
| 13:31 |
csharp |
JBoyer: looks like that does it' |
| 13:34 |
csharp |
yep - confirmed on multiple searches |
| 13:34 |
Dyrcona |
Well, that's one less thing to worry about. JBoyer++ |
| 13:34 |
csharp |
I'm still going to report a bug |
| 13:35 |
Dyrcona |
Not sure we can do anything about it. |
| 13:35 |
csharp |
hmm |
| 13:42 |
JBoyer |
I'm not sure how feasible it is but perhaps if clients know they're part of an apache process that's closing they could dash off a message to the router saying "forward my open requests to /dev/null" and that could be logged differently / not at all. |
| 13:42 |
JBoyer |
Not I'm not actually sure how to do that without just doing jazz-hands there in the middle. |
| 13:42 |
JBoyer |
and saying the words "magic happens here" |
| 13:47 |
|
mantis1 left #evergreen |
| 13:48 |
|
mantis1 joined #evergreen |
| 15:07 |
|
jihpringle joined #evergreen |
| 16:36 |
|
mantis1 left #evergreen |
| 16:49 |
|
rjackson_isl_hom joined #evergreen |
| 16:51 |
|
laurie joined #evergreen |
| 17:14 |
|
mmorgan left #evergreen |
| 17:32 |
|
sandbergja joined #evergreen |
| 17:56 |
|
jihpringle joined #evergreen |
| 18:01 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 22:29 |
|
sandbergja joined #evergreen |
| 23:28 |
|
Cocopuff2018 joined #evergreen |
| 23:35 |
|
sandbergja joined #evergreen |
