Time
S
Nick
Message
14:16 donsizemore joined #dataverse
15:00 pdurbin joined #dataverse
15:49 poikilotherm2
Hey pdurbin :-)
16:06 pdurbin
poikilotherm2: hi
17:20 nightowl313 joined #dataverse
17:21 nightowl313
sorry folks got another question ... we have a researcher who is uploading a very large number of files into a dataset ... and some of them are also going to be zipped, so very large files
17:21 nightowl313
we have direct upload set up on a store so that we can choose that if the dataset will have large files... but in this case he will have both
17:22 nightowl313
in the case where a dataset will have both large and small files, do we just choose the store that is using direct upload? or go back and forth ... a little confused about that
17:22 nightowl313
it is the same bucket
17:22 nightowl313
but just has two different storage locations configured in dataverse
17:23 pdurbin
In the GUI you can switch between the two stores, right?
17:23 nightowl313
yes ... so do they just switch between the two depending on the type of files being uploaded?
17:23 nightowl313
back and forth?
17:24 pdurbin
sounds suboptimal :)
17:24 pdurbin
I'd suggest creating an issue and telling your story.
17:24 nightowl313
just curious how others are handling this ... do most datasets fall into one or the other category?
17:25 pdurbin
Maybe there could be a threshold for using one store or another?
17:25 pdurbin
If you don't get an answer here you could also ask on the mailing list.
17:26 pdurbin
As a developer I only use the one store and I only upload small files.
17:27 nightowl313
yes, i will send a community email ... thanks
17:27 pdurbin
sure
17:33 poikilotherm2
pdurbin if you have a minute, maybe you can tell me if API tests already cover ingest enough to make sure I don't break anything with https://github.com/IQSS/dataverse/pull/7613
17:34 pdurbin
The tests look good to me: https://jenkins.dataverse.org/job/IQSS-Dataverse-Develop-PR/job/PR-7613/3/testReport/
17:35 pdurbin
And yes, ingest is tested. See, for example, UtilIT.MAXIMUM_INGEST_LOCK_DURATION.
19:21 poikilotherm2
Nice nice nice pdurbin.
19:22 pdurbin
yay tests
19:24 poikilotherm2
If you feel like it, please give that PR a pat and move it along 😎 Fixing logs is a good thing I assume...
19:27 pdurbin
I'm kind of heads down on disable user.
19:29 pdurbin
poikilotherm2: how will it affect OIDC?
19:30 pdurbin
When does your stuff call session.setUser?
19:30 poikilotherm2
Please request a review and mention me, I'll do a deep dive tomorrow when sitting in front of a screen. Road warrior right now.
19:30 poikilotherm2
Ah ah ah I don't know from the top of my head...
19:31 poikilotherm2
Let me do a quick check
19:31 pdurbin
No rush. Watch the road!
19:31 pdurbin
Toonces, look out!
19:35 poikilotherm2
Hmm I'm not sure I can find this right now...
19:35 pdurbin
No worries. It's on my list. I'll remind you. :)
19:36 poikilotherm2
Isn't the flow like the user object is retrieved at some other place?
19:36 poikilotherm2
There is all the getUserRecord() things
19:36 poikilotherm2
Should we look into this before or after call tomorrow?
19:37 poikilotherm2
I can refamiliarize with the auth flow and codebase
19:37 pdurbin
I forget if you build on top of the OAuth code or not.
19:38 poikilotherm2
I do! I reused the interfaces etc, which is why the OIDC stuff is a subpackage
19:38 pdurbin
Ok, so maybe I'm already testing that code path.
19:39 poikilotherm2
Definitely good chance
19:39 pdurbin
OIDCAuthProvider extends AbstractOAuth2AuthenticationProvider
19:39 pdurbin
good
19:43 poikilotherm2
About an hour to go till Mars rover landing...
19:45 pdurbin
Oh! Nice. I remember the one from when my kids were small.
22:06 pdurbin left #dataverse
