| Time |
Nick |
Message |
| 00:30 |
|
tsadok joined #evergreen |
| 01:17 |
|
dmoore joined #evergreen |
| 07:53 |
|
BDorsey joined #evergreen |
| 08:00 |
|
Dyrcona joined #evergreen |
| 08:09 |
|
collum joined #evergreen |
| 08:13 |
Dyrcona |
Bmagic++ # I hope he got some sleep last night. |
| 08:27 |
|
mantis1 joined #evergreen |
| 08:58 |
|
dguarrac joined #evergreen |
| 09:06 |
Dyrcona |
Ingest seems to be way too slow on a 3.10 database. I'll have to check Lp for any magic sauce. I'll also have to make sure that the symspell triggers were not reinstated at some point on this db. |
| 09:07 |
Dyrcona |
ps -o etime on the initial process says: 6-18:20:53. |
| 09:08 |
Dyrcona |
It's about 16% of the way through. |
| 09:26 |
Dyrcona |
symspell triggers were disabled.... |
| 09:33 |
Dyrcona |
Huh. I can't really tell because I don't have good timing data, but it seems to be moving along now... All I did was shutdown osrf services on a couple of virtual machines and install updates. Nothing was rebooted or whatever. |
| 09:34 |
Dyrcona |
Maybe the mostly idle connections were slowing things down? |
| 09:35 |
|
stephengwills joined #evergreen |
| 09:40 |
Dyrcona |
Doing a rough calculation based on batches/minute, the speed increase is roughly 13% from 0.53 to 0.6. |
| 09:42 |
Dyrcona |
Nope. I stand corrected 73% increase. I had an error in my formula to calculate batches/minute for the 3636 batches that had been initially processed. |
| 09:42 |
|
mmorgan joined #evergreen |
| 09:43 |
Dyrcona |
Of course, half an hour isn't much data for a comparison. I've restarted services on 1 vm, and I'll check again in a few hours. |
| 09:45 |
Dyrcona |
I wonder, too, if it matters which database is being used for services when you have more than one. I can see a case where doing an ingest on database A while services are communicating with database B could slow the ingest by keeping parts of B in memory and making that memory unavailable to A. |
| 09:47 |
Dyrcona |
Memory use on the database server does not seem to have changed. |
| 09:48 |
mantis1 |
Trying to develop a list in the patron summary that shows holds notification preferences based on user settings |
| 09:48 |
mantis1 |
I tried many iterations of macros to get it to show but nothing seems to reflect |
| 09:48 |
mantis1 |
one example: <li ng-show="hold_notify_type.email() == 't'">[% l('Email') %]</li> |
| 09:49 |
mantis1 |
I'm not sure if the hold_notify_type.email is the right one but looking for ideas |
| 09:51 |
Dyrcona |
mantis1: If the changes are in a tt2 file, then I would do it all with tt2 directives rather than mixing JS and tt2. |
| 09:51 |
mantis1 |
ok that's fair enough |
| 09:51 |
mantis1 |
though I did see the same syntax in the file that refers to another tt2 for the privacy waiver |
| 09:53 |
Dyrcona |
Are you adding something to an existing tt2? If that's the case it might be useful to make a separate file and include or process it. |
| 09:53 |
mantis1 |
yes it;s for the t_summary file |
| 09:53 |
Dyrcona |
You'll also need to make sure that the user settings data is included. |
| 09:53 |
* Dyrcona |
has a look. |
| 09:54 |
|
jvwoolf joined #evergreen |
| 09:56 |
Dyrcona |
I see. You might not be able to do what I suggested about doing it all in tt2 directives.... |
| 09:56 |
mantis1 |
right I was more or less copying the syntax from the privacy waiver |
| 09:57 |
mantis1 |
We're trying to add another line that shows if a patron has notifications for Phone, email, sms |
| 09:57 |
mantis1 |
so I was trying to go off of holds_notify_type |
| 09:58 |
Dyrcona |
Well, you might have to modify the circ patron service or circ/patron/app to retrieve user settings. I had a quick look and I didn't see them being retrieved. I probably missed something, though. |
| 09:59 |
mantis1 |
ok that gives me an idea of where to go |
| 09:59 |
mantis1 |
Dyrcona++ |
| 10:00 |
Dyrcona |
That's under Open-iLS/web/js/ui/default/staff... for AngularJS. You'd have to do something entirely different for Angular. |
| 10:00 |
Dyrcona |
That last comment is mainly for the logs. In case someone else goes searching for something similar. |
| 10:01 |
mantis1 |
regctl.js? |
| 10:03 |
Dyrcona |
I looked at circ/patron/app.js and circ/services/patron.js. If this is for patron edit, then regctl.js might be the place to look. |
| 10:03 |
mantis1 |
Dyrcona++ |
| 10:03 |
Dyrcona |
I didn't look hard enough to see what is being used by that interface. |
| 10:04 |
mantis1 |
I figure that the user settings would pull from this file |
| 10:04 |
mantis1 |
I did have it open for reference |
| 10:05 |
Dyrcona |
Yeah regctl.js pulls user settings, but I don't think the summary uses it. You could use it as a model for the changes you need to make. |
| 10:06 |
Dyrcona |
Related to my previous ramblings, the ingest seems to be maintaining the 73% performance boost after an hour of running with services from only 1 vm going. |
| 10:13 |
|
rfrasur joined #evergreen |
| 10:24 |
Dyrcona |
It also happens to be the VM talking to the other database that's running services, so I guess that doesn't matter. |
| 11:31 |
mmorgan |
mantis1++ # Great idea! |
| 12:11 |
|
jihpringle joined #evergreen |
| 12:18 |
|
Christineb joined #evergreen |
| 13:34 |
jeff |
@decide deduplicating bibs or deduplicating patrons |
| 13:34 |
pinesol |
jeff: That's a tough one... |
| 13:34 |
jeff |
yep. |
| 13:36 |
Dyrcona |
Patrons is probably tougher, to be honest. :) |
| 13:37 |
Dyrcona |
"You need the Doofenschmirtz Evil, Incorporated Dedupinator 3000(tm)!" |
| 13:38 |
Dyrcona |
Sorry. Flashbacks to my daughter's childhood. ;) |
| 14:24 |
|
mantis1 left #evergreen |
| 15:20 |
|
kworstell-isl joined #evergreen |
| 17:12 |
|
mmorgan left #evergreen |
| 17:24 |
|
jvwoolf left #evergreen |
| 18:04 |
|
jihpringle joined #evergreen |
| 21:15 |
|
stephengwills left #evergreen |
