[Archivesspace_Users_Group] Error trying to load via spredsheet

Robert T. Wilson rtwilson at cornell.edu
Wed May 10 11:12:57 EDT 2023


Blake, one other question. Background job links in 3.3.1 all point to localhost instead of domain http://localhost:8080/jobs/208/file/309?ext=.xlsx for example. I had setup a few instances before migrating prod, and these all reflected that. I was wondering if this was a configuration issue on my end and possibly related. If configuration issue, I'm all out of ideas on what could be causing it. Links in 2.5 with same setup as my original 3.3.1. test box show the domain url instead.


Robert


________________________________
From: archivesspace_users_group-bounces at lyralists.lyrasis.org <archivesspace_users_group-bounces at lyralists.lyrasis.org> on behalf of Blake Carver <blake.carver at lyrasis.org>
Sent: Wednesday, May 10, 2023 09:10
To: Archivesspace Users Group <archivesspace_users_group at lyralists.lyrasis.org>
Subject: Re: [Archivesspace_Users_Group] Error trying to load via spredsheet


It looks like you're hitting a known issue, with no known fix yet. Here's the JIRA with some details:

https://archivesspace.atlassian.net/browse/ANW-1427



Long story short, the importer just falls apart sometimes.  If you can replicate this problem reliably, be sure to comment there.



All I can say is SOMETHING goes wrong, and then it just stops working.


Restarting always clears it up.



________________________________
From: archivesspace_users_group-bounces at lyralists.lyrasis.org <archivesspace_users_group-bounces at lyralists.lyrasis.org> on behalf of Robert T. Wilson <rtwilson at cornell.edu>
Sent: Wednesday, May 10, 2023 9:58 AM
To: archivesspace_users_group at lyralists.lyrasis.org <archivesspace_users_group at lyralists.lyrasis.org>
Subject: [Archivesspace_Users_Group] Error trying to load via spredsheet

Hi folks,

We've just moved to 3.3.1, and I am seeing an odd issue when users try to load via spreadsheet. We can select our file, but when we click "import selected" we get the generic "something went wrong" 500 page.

I've attached debug logs, but these were logs at warn level

F, [2023-05-09T17:59:29.110284 #23230] FATAL -- :
F, [2023-05-09T17:59:29.110478 #23230] FATAL -- : RuntimeError (Couldn't make an ID out of URI: ):
F, [2023-05-09T17:59:29.110534 #23230] FATAL -- :
F, [2023-05-09T17:59:29.110576 #23230] FATAL -- : app/controllers/jobs_controller.rb:51:in `create'
app/controllers/application_controller.rb:744:in `set_locale'
E, [2023-05-09T17:59:31.130959 #23230] ERROR -- : Thread-8720: Unhandled exception!
E, [2023-05-09T17:59:31.131241 #23230] ERROR -- :
invalid byte sequence in UTF-8

If a user is seeing this, I am easily able to reproduce. With the invalid byte sequence message, I assumed it was something in the spreadsheet (attached is associated with debug log), BUT the workaround I've found makes me think this is something else. The workaround: If I try creating another background job for resource in question like generate pdf then I try the load via spreadsheet again, the job queues and runs without issue or delay.

Has anyone seen something like this or have any recommendations?

Thanks in advance for any assistance y'all can provide.

Robert

Robert T. Wilson (he/him)
Cloud Systems Administrator
Cornell University Library
rtwilson at cornell.edu
607.254.3483
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lyralists.lyrasis.org/pipermail/archivesspace_users_group/attachments/20230510/7ca45198/attachment.html>


More information about the Archivesspace_Users_Group mailing list