Fix improper Lucene write.lock handling in DNN Platform

[trilogy-ai]

Revisions to the indexing code ensure proper handling of Lucene's write.lock files. Modifications include introducing a more robust lock management strategy utilizing Lucene's LockFactory and implementing retry logic with exponential backoff during lock acquisition attempts to mitigate temporary lock contention.

Execution ID: 01HQQC81Z0Q62GC92P3JS65DPC

[valadas]

This is new, can we have some info about what is @trilogy-ai ?

[mitchelsellers]

@valadas I'm looking into this

[Tantalon]

Please feel free to close this.

trilogy-ai is a bot that connects our work management system to AI agents so that work can be completed automatically. For code fixes, it generates a ticket like this for Sweep AI to handle.

We only intended it to run on repositories in the trilogy-group organization. In this case, because there are engineering maintenance requests for DNN, it created this ticket. We've added a check to trilogy-ai, so no more tickets should be created. Apologies for the noise.

[valadas]

Will a PR be coming to fix this in Platform or is this something Evoq specific ?

[Tantalon]

An engineer is working on a Platform fix for it.

[valadas]

Awesome, then I think we can keep this issue here