After the recent outage was resolved (for the file-search), I noticed issues with time retrieval, token usage, and the number of queries sent to the vector store. Previously, the process was very fast and did not have any problems with the number of queries being sent.
Is anyone else facing this issue?