Disk I/O Errors Caused by Large ReviewBot Cache File in Review Board

[Rashmi Sherawat]

Hi,

We have integrated ReviewBot using Claude with Review Board, and it is now being used by a growing number of users across our organization. Currently, all AI review requests are processed through a single account ID.

We are encountering an issue where the worker machines disconnect from the Review Board ReviewBot extensions due to a “Disk I/O error on the worker machines caused by a large rbt cache file.”

To mitigate this, we have been manually clearing the cache using the command with id with which it is running:

rbt clear-cache

This removes the file:

~/.cache/rbtools/apicache.db

We have also added this command to a cron job that runs every hour. Despite this, the error still occurs intermittently. 
Where the apicache.db file generated is having ~500GB space already and home area of id with which it is running is also having ~10GB space.

If disk space is not the primary issue, could you please help explain why this error is occurring and suggest possible solutions or best practices to prevent it?

Thanks

Rashmi

[Rashmi Sherawat]

Any help on thee below issue please?

[Christian Hammond]

Hi Rashmi,

We're going to be looking into this as part of the work we have planned for both Review Bot and RBTools. We can work with you to get expedited custom patches for your setup under a support contract. If you'd like to discuss this, I can reach out to you directly with what that would look like and get that process going.

Christian

--

---
You received this message because you are subscribed to the Google Groups "Review Board Development" group.
To unsubscribe from this group and stop receiving emails from it, send an email to reviewboard-d...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/reviewboard-dev/41c57019-942e-4cc7-a0a1-680e2ec38c05n%40googlegroups.com.

--

Christian Hammond

President/CEO of Beanbag

Makers of Review Board