| Time |
Nick |
Message |
| 07:39 |
|
redavis joined #evergreen |
| 07:57 |
|
cbrown-isl joined #evergreen |
| 08:06 |
|
BDorsey joined #evergreen |
| 08:41 |
|
mmorgan joined #evergreen |
| 08:45 |
|
Dyrcona joined #evergreen |
| 09:05 |
|
mmorgan1 joined #evergreen |
| 09:17 |
|
dguarrac joined #evergreen |
| 09:25 |
|
mantis joined #evergreen |
| 09:51 |
|
sandbergja joined #evergreen |
| 09:54 |
sandbergja |
Dyrcona: eeevil: you both said you'd be interested in taking a look at a pullrequest making some changes to the release process (git tags instead of "tag branches", etc.). I've got a pullrequest up on bug 2060734 -- if you have the time to take a look, I'd appreciate your feedback! |
| 09:54 |
pinesol |
Launchpad bug 2060734 in Evergreen "Various changes to the release process" [Undecided,New] https://launchpad.net/bugs/2060734 |
| 10:14 |
|
collum joined #evergreen |
| 10:28 |
csharp_ |
sandbergja++ |
| 10:31 |
|
kworstell-isl joined #evergreen |
| 10:41 |
|
kworstell-isl joined #evergreen |
| 11:25 |
|
Christineb joined #evergreen |
| 12:43 |
mantis |
When manually testing hold expiring soon notices, I have a test set up so that the hold that is captured has a shelf expiration two days from now, but when I run a script, nothing runs. Not sure if anyone has any ideas or tried testing this notice before. |
| 12:52 |
berick |
mantis: we use it w/ delay = '-2 days' and max_delay = '-1 day' |
| 12:53 |
mantis |
ah ok |
| 12:53 |
mantis |
I have a max delay of 2 days |
| 12:53 |
mantis |
I can give that a shot |
| 12:53 |
* Dyrcona |
was typing something long, but yeah, you want -2 days. |
| 12:54 |
Dyrcona |
Negative for before, positive for after. |
| 12:54 |
eeevil |
mantis: think of delay and max_delay as the age of the timestamp in the field pointed at by delay_field (in fact, it's exactly that!) |
| 12:54 |
eeevil |
as in, the output of the age() function in the db |
| 12:55 |
mantis |
eevil: I also thought had to do something with the timestamp, but testing wasn't as fruitful |
| 12:55 |
mantis |
and when you edit the hold dates in the regular circ module, it doesn't give a timestamp option |
| 12:55 |
mantis |
berick: I gave that a shot and no dice |
| 12:56 |
mantis |
my shelf expire time is set to 6-12-2024 00:00:00-04 |
| 13:01 |
Dyrcona |
mantish: action_trigger.event_hook specifies a "core_type" for the event. This is the main IDL class used by the event. action_trigger.event_definition specifies a delay_field, delay, and max_delay. |
| 13:01 |
Dyrcona |
The delay and max_delay (if present) are added to the value of the delay_field of the object being processed to determine if an event should be generated for that object. |
| 13:03 |
Dyrcona |
To process things from the past, i.e. a 30-day overdue event, you would use positive values in delay and max_delay. For future things, like hold expires in 2 days, use negative values. So a delay of '-2 days' and a max_delay of '-1 days' would get everything between tomorrow and two days from now. |
| 13:03 |
Dyrcona |
mantis^^ |
| 13:04 |
Dyrcona |
If my fingers would learn to cooperate with my brain, that would be great. |
| 13:06 |
Dyrcona |
Also, if you only set the delay field, then it does exactly two days from now, which is probably not what you want. It's always good to specify a max_delay as well in order to have a range. |
| 13:06 |
Dyrcona |
The action trigger process is doing timestamp math, not date math, so the hours matter. |
| 13:08 |
Dyrcona |
Maybe you know all that, but I guess it won't hurt for the logs to know, too. :) |
| 13:10 |
mantis |
Thanks for the explanation; I guess we could have just kept it as it was |
| 13:10 |
mantis |
I'm just confused as to why nothing is coming back at all |
| 13:10 |
berick |
i often have to check the SQL logs to make sure my date math is right on these |
| 13:11 |
mantis |
I'll play with the fields a bit |
| 13:11 |
mantis |
berick++ |
| 13:11 |
mantis |
eevil++ |
| 13:11 |
mantis |
Dyrcona++ |
| 13:11 |
Dyrcona |
I'd check the action_trigger_filers.json in /openils/conf, or if you're using custom filters, make sure that file is there. |
| 13:11 |
* Dyrcona |
often forgets the filters when testing triggers. |
| 13:15 |
Rogan |
This is also sent out on the general dev lists but details for the 2024 Hack-A-Way are now available: https://wiki.evergreen-ils.org/doku.php?id=hack-a-way:hack-a-way-2024 |
| 13:48 |
|
terranm joined #evergreen |
| 13:54 |
terranm |
Claiming 1419 |
| 13:58 |
pinesol |
News from commits: LP#2003090 Angular Daily Cash Report Missing Print Capability <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=b18e60fd19211310e6200f65b101c5056871577c> |
| 14:06 |
csharp_ |
@decide rebase or freebase |
| 14:06 |
pinesol |
csharp_: go with freebase |
| 14:28 |
pinesol |
News from commits: LP1788063 Manage Copy Alerts Wording Change <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=3e27aee68a53cb0f07ed0cdcf2ab115d2cb95dc1> |
| 14:28 |
pinesol |
News from commits: LP#1783408 Stamp upgrade script <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=0e8ffc115fa55bed9dc4c7c7668d647e844912bd> |
| 14:28 |
pinesol |
News from commits: LP#1783408 "Disable Highlighting" Box Should be sticky <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=f066d68c3ecd3d0dcde676a4e4674ae0be342b73> |
| 14:28 |
pinesol |
News from commits: LP#1970946 Staff Catalog Option to Disable Highlighting <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=afdefb94aeef3f2f68acca0fdb5a13f408ca3dde> |
| 14:58 |
pinesol |
News from commits: LP2015112 Hide Edit Call Number Link if Missing Permissions <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=5a0c69488ed49da333a2f337833d940b9a6c6c57> |
| 15:59 |
pinesol |
News from commits: LP#2068755 - Hold Shelf Columns : Add User Permission Group <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=ab74505a43fc121f5a3f6f9675742fb4ca54019f> |
| 15:59 |
pinesol |
News from commits: LP2058747 Fix "(current)" in staff catalog pagination <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=d678b450e17d5087f916d0755d2d6255e92aaafb> |
| 15:59 |
pinesol |
News from commits: LP1861331 Configurable name for row flair column <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=000957cbc38ca9bd82dbf2923727a0919b6e0d72> |
| 16:29 |
pinesol |
News from commits: LP2058287 Label for Missing Pieces barcode input <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=65766486d77dae38d169d9f5610d03e026418623> |
| 16:56 |
|
mantis left #evergreen |
| 17:02 |
|
mmorgan1 left #evergreen |
| 17:03 |
|
JonHGeorg joined #evergreen |
| 17:11 |
JonHGeorg |
Is the IRC channel dead? I ask because I thought I remembered something about that being discussed at the conference and the link at the top for the logs doesn't work anymore. |
| 17:16 |
berick |
JonHGeorg: we're still here. |
| 17:16 |
berick |
today's logs http://irc.evergreen-ils.org/evergreen/2024-06-10 |
| 17:35 |
JonHGeorg |
Great to hear. My browser doesn't like that link either. Shrug. |
| 17:48 |
JonHGeorg |
So question. We're having issues with action_trigger.pl. Holds and auto-renewals aren't going out. Emails are still going out for receipts, etc so it's no an email issue. We had a space issue on our log server a while back, and then shortly after that this cropped up. There is a /tmp/action-trigger-LOCK dated 6/2, which happens to be the last date |
| 17:48 |
JonHGeorg |
that there are invalid events in action_trigger.event. I'd queried that table trying to find errors, pending status, etc and on a whim ran a search against invalid entries. They all stopped after 6/2 which I felt was probable grounds indicating something crashed. My question is, is fixing the issue as simple as killing the PIDs related to |
| 17:48 |
JonHGeorg |
action_trigger.pl, deleting the lock file, and restarting the cron job manually? Or will that take other things down? When I do a tail -f /var/log/syslog|grep 'action_trigger' I get a ton of entries indicating part of it is still running, but they all appear to be search related. Thanks |
| 19:10 |
|
JonHGeorg joined #evergreen |
| 21:26 |
Bmagic |
JonHGeorg: first step is to see if the action_trigger process is running: ps aux|grep action_trigger. If it's not running and* you still have a LOCK file, then it's safe to delete the LOCK file. That might* be the whole problem. The process won't start if it finds a LOCK file for itself |
| 22:41 |
JonHGeorg |
Thank you. It was listed as running. Chris at Georgia Pines ended up providing a lot of help to me today. One thing that helped me was querying action_trigger.event to see what was happening today after getting a complete list of everything from action_trigger.event_definition so I could see which ones were erroring out, etc. I was getting zero |
| 22:41 |
JonHGeorg |
results for today and had been since the lock file's date. So I killed the PIDs, removed the lock file, manually ran action_trigger and I already see completed notifications populating action_trigger.event. This is a modification to a query he gave me. select * from action_trigger.event where state in ('complete') and date(start_time) = |
| 22:41 |
JonHGeorg |
'yyyy-mm-dd'; --today's date |
| 22:41 |
JonHGeorg |
If it had been going on longer it would have made sense to follow his advice of adjusting max_delay but Beth and I didn't think it was needed in this instance, although wildly helpful to know nonetheless. Thank you everyone for your help. |
| 23:41 |
Bmagic |
ok great! |
