Bulk Deletion of Old Recordings

[Savneel Kant]

Hi,

 

We currently have a substantial number of recordings, spanning thousands, that require deletion. These recordings date back to 2022 and 2021. Unfortunately, we've encountered difficulties in removing them using our standard retract and delete method.

On a positive note, we've been successful in retracting and deleting recordings from 2023 and 2024 using Opencast version 8.5.

We're reaching out to inquire if there's a method to efficiently remove all recordings from 2021 to 2022, including any associated links from the presentation server, in bulk from the database, or through an alternative solution.

Any assistance or guidance you can provide on this matter would be greatly appreciated.

[Katrin Ihler]

Hi Savneel,

what problems do you run into with those old recordings? Maybe a publication isn't properly retracted.

Also Opencast 8 is long out of date, which means solutions we suggest might not be supported by your version, and it might contain bugs contributing to your problem.

If all else fails, you could scrub all that data from the different places (database, NFS share, then rebuilding the indexes), but it's easy to miss a spot. Better to get the standard delete behavior working. Then you could bulk delete by writing a small script to call the related rest endpoint.

Cheers,

Katrin

To unsubscribe from this group and stop receiving emails from it, send an email to users+un...@opencast.org.

-- 
ELAN e.V.
Karlstr. 23
D-26123 Oldeburg

elan-ev.de

[Savneel Kant]

Hi Katrin,

Thank you for your email.

When we try to delete an old recording we get the following error:

Attached is the log file. This was generated when I tried to delete this single old recording. 

Am not so experienced with this setup, as this was installed and configured by a previous colleague who has left the organization. Would you be able to direct me to some notes or guides on how the following can be done as per your message:

"scrub all that data from the different places (database, NFS share, then rebuilding the indexes),"

Regards

Savneel Kant

[Katrin Ihler]

Hi Savneel,

it blows up here:

User workflowCreator = userDirectoryService.loadUser(workflow.getCreatorName());
String workflowOrgId = workflowCreator.getOrganization().getId();

Looks like the user that created the workflow belonging to the event no longer exists. This is a bug that was most likely fixed by https://github.com/opencast/opencast/pull/1639 

And the funny thing is, you actually only have to move one minor version up to get that fix - to 8.6! 😂

Cheers,

Katrin

[Savneel Kant]

Hi Katrin,

Thank you so much for the help. It helped to sort out the issue. :) 

Regards

Savneel Kant

[Katrin Ihler]

Awesome! :D

I'd still recommend upgrading to a currently supported version though, so you don't stumble over bugs that are already fixed.

Katrin

[Savneel Kant]

thank you for the recommendation Katrin. I have installed the version 14.10 on RHEL. Really appreciate your guidance. 

Regards

Savneel Kant