| Time |
Nick |
Message |
| 03:53 |
|
beardicus7 joined #evergreen |
| 08:59 |
|
Dyrcona joined #evergreen |
| 09:01 |
|
mmorgan joined #evergreen |
| 10:34 |
|
Christineb joined #evergreen |
| 11:43 |
|
jihpringle joined #evergreen |
| 11:53 |
goood |
Bmagic: re batch edit, yes, that's why the button isn't on the bucket content interface -- it would suggest one could batch edit just the selected set, which the batch update template builder UI can't facilitate today. now, that said, the batch update backend module accepts multiple bib IDs via repeated use of the the recid CGI param, so the backend could support selected-bib update, but the Angular batch update template builder needs to learn how to |
| 11:53 |
goood |
accept a list of IDs to pass along in that way (or, we could make the batch editor backend parse a comma-separated list, or have the Angular generate a file via the Blob JS interface, or ... something). it's non-trivial, but doesn't require all new code to do it. |
| 11:54 |
Bmagic |
gotcha, thanks for the clairification |
| 11:59 |
goood |
Dyrcona: I'll be merging those 2 API bugs today (hopefully). thanks! the is_rendered_success_type() fix in particular is great |
| 11:59 |
Dyrcona |
goood++ |
| 12:24 |
|
kmlussier joined #evergreen |
| 14:24 |
Dyrcona |
Wouldn't you know it. The server ran out of RAM and killed PostgreSQL while processing the last 4 batches. |
| 14:25 |
Dyrcona |
3553 of 3557 processed. |
| 14:36 |
Dyrcona |
Bmagic: you around? |
| 14:36 |
Bmagic |
sure am |
| 14:37 |
Dyrcona |
Since dev ran out of memory and killed postgresql again, I'm thinking about setting overcommit_memory to 2 (which means don't overcommit), even though I think that might have caused a crash on Tuesday. What do you think about that? |
| 14:39 |
Dyrcona |
PostgreSQL recommends 2, but general applications are probably best with 0, and Redis likes 1 apparently. (But we're not using Redis here, yet.) |
| 14:40 |
Dyrcona |
For high transactions web servers, I've also seen setting it to 1 recommended. |
| 14:41 |
Dyrcona |
Maybe I'll go with 0 and let the kernel figure it out. |
| 14:42 |
Dyrcona |
I don't mind having this discussion in public since it might help others with their development environments. |
| 14:44 |
* Dyrcona |
tries 0 (the default). |
| 14:47 |
Dyrcona |
I wonder if these differences between recommendations for Redis and PostgreSQL would contribute to issues running them both on the same machine? |
| 14:47 |
Dyrcona |
Might explain some of the things we've seen. |
| 15:00 |
Dyrcona |
And, contrary to what I said in an email to Bmagic, I am not going to restart the pingest. I'm going to leave it alone for now and wait to see if there are issues next week. It's probably not even necessary going from 3.15 to 3.17. |
| 16:04 |
Dyrcona |
Have a good weekend, everyone! |
| 16:25 |
goood |
grabbing 1522 |
| 17:14 |
|
mmorgan left #evergreen |
| 19:12 |
|
kmlussier left #evergreen |
