jasig-ssp IRC Logs-2013-03-22

Owned by Apereo Infrastructure
Last updated: Mar 22, 2013
4 min read

[09:55:51 CDT(-0500)] <TonyUnicon> is Unicon's jabber server down?

[09:56:45 CDT(-0500)] <dmccallum54> i'm connected and chatting w/ greg

[09:57:30 CDT(-0500)] <TonyUnicon> back on now, strange

[10:01:53 CDT(-0500)] <TonyUnicon> there was no issues really with Jackson and unmarshalling arrays

[10:02:11 CDT(-0500)] <dmccallum54> Jackson's nice

[10:02:13 CDT(-0500)] <TonyUnicon> not as fragile as I thought

[10:04:13 CDT(-0500)] <TonyUnicon> I wasn't so bold as to try to unmarshall the javascript array to a java list

[10:04:23 CDT(-0500)] <TonyUnicon> do you think that is just as painless?

[10:04:52 CDT(-0500)] <dmccallum54> i do, actually

[10:05:08 CDT(-0500)] <TonyUnicon> ill give it a shot next build

[10:06:43 CDT(-0500)] <TonyUnicon> so I was able to do the batch update with HQL fairly easily

[10:07:00 CDT(-0500)] <TonyUnicon> the HQL is an update with an in clause

[10:07:33 CDT(-0500)] <TonyUnicon> the only thing about that is that with large numbers

[10:07:34 CDT(-0500)] <TonyUnicon> or large number of elements in an 'in' clause

[10:07:34 CDT(-0500)] <TonyUnicon> and cause problems

[10:07:34 CDT(-0500)] <TonyUnicon> can cause problems*

[10:08:38 CDT(-0500)] <TonyUnicon> unless they plan on reassigning several hundred students at a time it should be ok

[10:09:04 CDT(-0500)] <dmccallum54> yeah, i think the limits are in the 1000s these days

[10:09:58 CDT(-0500)] <dmccallum54> tho is oracle still the grumpy one here?

[10:10:08 CDT(-0500)] <TonyUnicon> heh yeah

[10:10:15 CDT(-0500)] <TonyUnicon> oracle is where ive had that problem

[10:13:52 CDT(-0500)] <dmccallum54> so

[10:14:09 CDT(-0500)] <dmccallum54> let's ask jason on the call about how many students would typically be reassigned at once

[10:15:01 CDT(-0500)] <dmccallum54> my guess is the thing to do would be to just open a nice-to-have ticket for batching these updates so we don't forget about the issue, but treat it as low priority, at least until we look at oracle support

[10:16:00 CDT(-0500)] <TonyUnicon> it is not hard to break it up

[10:16:10 CDT(-0500)] <TonyUnicon> I can just do it

[10:16:15 CDT(-0500)] <dmccallum54> heh. ok

[10:16:19 CDT(-0500)] <TonyUnicon> in batches of 300 or 500 hundreds to be safe

[10:16:31 CDT(-0500)] <TonyUnicon> there is no real penalty

[10:16:50 CDT(-0500)] <dmccallum54> a global config that sets the max 'in' batch size along w/ your other db config?

[10:17:05 CDT(-0500)] <TonyUnicon> yeah I can do that

[10:17:09 CDT(-0500)] <dmccallum54> k

[10:17:16 CDT(-0500)] <TonyUnicon> makes sense

[10:17:27 CDT(-0500)] <TonyUnicon> for future batching operations

[10:19:16 CDT(-0500)] <TonyUnicon> i'll also put a default batch size in one of the base daos, so the config is not required

[10:21:41 CDT(-0500)] <dmccallum54> that's fine. tho there's defaulting kinda built in w/ the property file loading mechanism

[10:22:04 CDT(-0500)] <dmccallum54> it will always fall back to the properties in the ssp-config.properties bundled in with the app

[10:22:19 CDT(-0500)] <dmccallum54> and you can inject those values with SpEL expressions

[10:23:55 CDT(-0500)] <dmccallum54> for an annotation-based version of that see PersonServiceImpl#perCoachSyncTransactions

[10:24:17 CDT(-0500)] <dmccallum54> there should be other XML based examples scattered around in the bean def files

[10:29:58 CDT(-0500)] <TonyUnicon> ok

[10:30:04 CDT(-0500)] <TonyUnicon> i'll do that then

[12:25:39 CDT(-0500)] <js70> Tony. look for test_data_generator.groovy.

[12:27:46 CDT(-0500)] <js70> Take a look at the file, you will need to set some of the parameters to generate the amount of students and coaches early alerts etc. you want. If you need to log in as one of the generated coaches you will need to run ant initdb in the uportal folder.

[12:29:07 CDT(-0500)] <js70> any questions, feel free to ask. There are still some bugs in how the data is generated (primarily in the journal entry side) but I am able to bring up students/ pull up data etc. I think I know what the issues are but if you find anything that is stopping you let me know and I'll see if I can fix it.

[12:30:59 CDT(-0500)] <TonyUnicon> how do you usually kick off this groovy file?

[12:31:07 CDT(-0500)] <TonyUnicon> its been awhile

[12:31:29 CDT(-0500)] <js70> I use the groovy plug in in eclipse. never used groovy in the wild.

[12:31:53 CDT(-0500)] <js70> its either a right click run as groovy script or you can add a configuration.

[12:32:20 CDT(-0500)] <TonyUnicon> alright, thank you for the info Jim

[12:32:25 CDT(-0500)] <js70> np

[13:10:20 CDT(-0500)] <dmccallum54> js70 can you create a new ticket for your ajax report form submit work?

[13:10:44 CDT(-0500)] <dmccallum54> that's really a separate thing altogether from 688, even tho the problem was originally reported there

[13:11:13 CDT(-0500)] <js70> yep. will do.

[13:11:19 CDT(-0500)] <dmccallum54> thx

[13:12:54 CDT(-0500)] <dmccallum54> our commit logs are going to be so focused!

[13:12:58 CDT(-0500)] <dmccallum54> it's very exciting

[13:13:48 CDT(-0500)] <js70> your weird:^)

[13:14:02 CDT(-0500)] <js70> you are weird. :^)

[13:18:49 CDT(-0500)] <js70> Just to be clear, I think weird is good thing and thanks for the acknowledgement for the team.

[13:24:53 CDT(-0500)] <dmccallum54> naturally

[13:26:18 CDT(-0500)] <js70> out for lunch.

[13:26:27 CDT(-0500)] <TonyUnicon> hah

[18:04:11 CDT(-0500)] <dmccallum54> js70 i'm having trouble merging c5218bd368c44394e9ee96373531ed700432ea11 into master. conflcts in DisabilityServicesReportControllerIntegrationTest.java. looks like there were recent changes https://github.com/Jasig/SSP/commit/f37a320a6bf9557f962c747a532b32ee3d82c4b8

[18:04:54 CDT(-0500)] <js70> k

[18:04:56 CDT(-0500)] <dmccallum54> is it reasonable to assume that the current state of that test class in master is correct? i.e. i can just discard the inbound diff from rel-1-2 and go with what's already in master?

[18:05:16 CDT(-0500)] <dmccallum54> looks like that patch was nearly a wholesale replacement

[18:05:17 CDT(-0500)] <js70> yes.

[18:05:21 CDT(-0500)] <dmccallum54> ok cool

[18:05:22 CDT(-0500)] <dmccallum54> thx