Resolved
Stabilized latencies.
Monitoring
The corrupted node have been drained and translation and username pod recreated.
Latencies are back to normal.
We continue to monitor.
Identified
A username pod shared on the same node have increased latency as well impacting username analysis
Identified
We are seing an increased latency on the analyze text endpoint.
A translation pod is in cause, rejecting most of the received requests.
