I am working on analysing a cloud run high latency issue. The app deployed in Cloud Run Service is developed in Java, Springboot, and Alloydb is the persistence layer. The app is one of the microservice and it is exposed a REST APi endpoint. The endpoint receives entries for adding new entry in Alloydb. The rate of receiving new entries via rest api is neither high nor low.
A particular day, unusually the cloud run service froze for 90 second. The cloud run logs show, db write has happened from the app which means app established database connection. Immediately after the db write , the service app froze almost 90 seconds before publishing next log. The java code does nothing except calling the repository method in springboot.
Cloud run logs show no spike or no heavy load. I need some clue on what could possibly have happened in that 90 seconds before resuming the service. Where can I search and investigate for next clue.
I do not have access to infrastructure level logs for the cloud run service like DNS logs, VPC Connector logs, PSC logs. IS there any place, I can look for clue.
The cloud run service is deployed in docker container. There is no special metnion of any thing in log regarding HikariCP. The log shoed db entry created successfully and 90 second silence, then after that, next log started to get printed.
I checked the database table, and the timestamp of the log (data entry created successfully) published exactly before the 92 second freeze and created timestamp is exactly same.