| Time |
Nick |
Message |
| 06:01 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 07:34 |
|
rjackson_isl_hom joined #evergreen |
| 07:46 |
|
collum joined #evergreen |
| 08:10 |
|
Dyrcona joined #evergreen |
| 08:11 |
|
rfrasur joined #evergreen |
| 08:13 |
Dyrcona |
So, my quip from last week that Clark needs better logging came from the reporter just crashing with nothing in any of the logs about it. It happened again last night/this morning. |
| 08:35 |
JBoyer |
I think the most common cause for the main clark process dying is any issue connecting to the db. Logging would be good, but connection retries would also be a big benefit to go along with. |
| 08:39 |
|
mmorgan joined #evergreen |
| 08:46 |
|
mantis joined #evergreen |
| 08:53 |
csharp_ |
most of the time for us it's the reports server filling up RAM and getting OOM killed |
| 08:54 |
csharp_ |
we have a server that just runs clark and connects to a DB replica that just does reports queries |
| 09:14 |
|
Keith-isl joined #evergreen |
| 09:22 |
Dyrcona |
csharp_: Our reports is similar, but the server also runs most cron jobs and hosts NFS. I don't see any OOM killer messages. In the past, I'd usually see these. It would also often leave a report in the currently_running view. That's not happening this time, so it could be something else. |
| 09:22 |
Dyrcona |
So far, I have found nothing useful or seemingly related in any of the logs. |
| 09:23 |
Dyrcona |
I've also dumped the schedule for last Thursday and today, and have not found any similarities, yet, between the reports that were running just before the two crashes, i.e. they look like very different sets of reports. |
| 09:33 |
Dyrcona |
Nothing in the PostgreSQL logs on db2, either. |
| 09:58 |
|
jvwoolf joined #evergreen |
| 10:09 |
Dyrcona |
I'm going to make a Lp to request that Clark use syslog instead of spewing to standard out/standard error. Daemons are actually supposed to close those descriptors, anyway. |
| 10:10 |
alynn26 |
Dyrcona++ just let me know and I will add all the heat possible. |
| 10:16 |
Dyrcona |
alynn26: bug 1933984 |
| 10:16 |
pinesol |
Launchpad bug 1933984 in Evergreen "Reporter Logging Improvements" [Undecided,New] https://launchpad.net/bugs/1933984 - Assigned to Jason Stephenson (jstephenson) |
| 10:16 |
Dyrcona |
alynn26++ JBoyer++ csharp_++ |
| 10:17 |
JBoyer |
Dyrcona++ |
| 10:52 |
|
dguarrac joined #evergreen |
| 11:10 |
|
Christineb joined #evergreen |
| 11:47 |
|
jihpringle joined #evergreen |
| 12:34 |
|
collum joined #evergreen |
| 12:52 |
|
collum joined #evergreen |
| 13:28 |
|
Keith_isl joined #evergreen |
| 14:07 |
|
rfrasur joined #evergreen |
| 14:46 |
|
Keith-isl joined #evergreen |
| 16:43 |
Bmagic |
How do people handle Authority delete files? |
| 16:45 |
rhamby |
like from backstage or marcive? |
| 16:48 |
rhamby |
I use a small opensrf script that checks agianst a bib file for deletions and matches against those in the system based on the 010, the resulting matches are usually small enough to delete by hand though it wouldn't take much to auto delete them |
| 17:15 |
|
mmorgan left #evergreen |
| 17:53 |
|
Stompro joined #evergreen |
| 18:02 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 22:59 |
|
Stompro joined #evergreen |
