Since a few hours we’re experiencing elevated latency and intermittent outages when accessing Piencone serverless in us-east-1. Happens to requests from instances on Vercel, Railway and locally. Should add that no changes were made on our side to infrastructure or implementation.
The dashboard and https://status.pinecone.io don’t show anything unusual.
This seems to affect /query endpoint only. With PINECONE_DEBUG=true there is nothing logged in the console when the endpoint is called and sometimes it resolves after 30 or 45 seconds, other times just time outs.
Does anyone else experience this?
// edit
After a few hours there are much less timeouts, maybe <5% of requests.
I noticed in our logs that we had this logged around 00:00 AM GMT:
// edit 2
The issue persists - and is not really timeout, just very long latency (eg. 5 min) that causes cascading timeouts.