How to prevent/deal with corrupted WAL for Azure Database for PostgreSQL flexible server.

Konrad Wölms 0

We are running a Postgres instance with a Timescale DB extension on Azure and are currently not able to log into the service from any client. With the error
"Connection refused Is the server running on that host and accepting TCP/IP connections?"

We had the same issue a month ago and attempted a restart of the instance, but the service was not able to fully restart and was basically stuck in "restarting". At the time we contacted the Support and their analysis yielded that somehow the WAL got corrupted and they reset it, presumably with something like "pg_resetwal" after which the instance was available again. We reset the database and afterwards the instance was running fine up until last Friday. While the error message does not provide enough details to be certain that this is again the same issue it seems likely.

I have the following related questions. Is there any way to figure out from the user perspective, whether WAL corruption is again responsible. If so, how to best diagnose what causes the WAL corruption in order to prevent in the future?

The instance belongs to a non-critical test system and we don't need the actual data. It is currently still left in the problematic state in order to better learn how to deal with this issue that now presumably happend a second time. However, we only have point in time restore available until this Thursday. We would like to use that to restore the database if we don't learn of better options by then. So any quick feedback is very much appreciated.

Mahesh Kurva 2,835 Reputation points Microsoft Vendor

2025-02-24T17:37:46.59+00:00
Hi @Konrad Wölms,

Welcome to Microsoft Q&A forum.

I'm sorry to hear about the issues you're facing with your Postgres instance. Let's try to address your concerns step-by-step.

The PostgreSQL logs can provide insights into whether WAL corruption is occurring. Look for error messages related to WAL files. You can find the logs in the pg_log directory of your PostgreSQL data directory.

This utility can provide information about the state of the database cluster. Running pg_controldata on your data directory can help you identify if there are any inconsistencies or issues with the WAL.

Sometimes, WAL corruption can be caused by underlying disk issues. Ensure that your storage is healthy and there are no disk errors.

Regularly back up your database to ensure you can restore it to a known good state if corruption occurs.

Use tools to monitor the health of your disks and storage systems. This can help you catch issues before they lead to corruption.

Always ensure that the database is properly shut down before restarting or performing maintenance. Abrupt shutdowns can lead to corruption.

Ensure that you are running the latest versions of PostgreSQL and TimescaleDB, as updates often include fixes for known issues.

I hope this information helps. Please do let us know if you have any further queries.
Adithya Prasad K 10 Reputation points Microsoft Vendor

2025-02-27T07:10:47.63+00:00

Hi Konrad Wölms
We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. In case if you have any resolution please do share that same with the community as it can be helpful to others. Otherwise, will respond with more details and we will try to help.
Adithya Prasad K 10 Reputation points Microsoft Vendor

2025-02-28T09:01:39.0933333+00:00

Hi Konrad Wölms,
We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. In case if you have any resolution please do share that same with the community as it can be helpful to others. Otherwise, will respond with more details and we will try to help.

Share via

How to prevent/deal with corrupted WAL for Azure Database for PostgreSQL flexible server.

Your answer