David Callies
David Callies
In HMA office hours, I think we aligned on starting with option 1. @turt2live - you are the only one who has mentioned they have bandwidth to give it a...
I like this idea, since S3 is likely going to be common. Is there a library we can use (boto3?) to help with configuration management?
Fun fact, this was the core implementation of HMA1.0, which was needed to hit our target scale! We used async lambda backed by SQS. @dougneal and I talked about this...
I'm leaving this one in the hold for now while we build out more of the feedback API calls on ThreatExchange/StopNCII/NCMEC. Some of them are quite tricky.
@juanmrad - I think you were actively working on this, so marking it to you, let me know if that's not sure.
New image is released, waiting for feedback from production deployments on whether this fixes the observed issues. Thanks @juanmrad for all the hard work here!
@prenner - any news on whether @juanmrad's changes solved the issue?
Discussed in office hours, @prenner is still seeing leaks in production with gunicorn even after PRs, so back to the drawing board
Yes - do you need them as issues or just to write them out? Steps: 1. Create a new file under pdq called `index2.py` which contains PDQSignalTypeIndex2, It can initially...
Not quite yet, we're missing the optimized solution - faiss IVF. IVF faiss indices should be used when the number of hashes are above some number (e.g. 1,000 hashes), and...