Phoenix app in a Docker container is filling log with this in production on an Azure Web App. Another post suggested hackney and httpoison is causing it. How can I find the cause of the errors? I can ssh to the container and run /app/bin/app_name remote to start iex. Do I need to create a ssh tu…

I did :sys.trace pid("0.277.3"), true to get this log from Debugging an unexpected process exit - #4 by hkrutzer I got the process id using Behold.process_by_memory from Elixir Processes: Observability - Samuel Mullen 2023-08-28T04:42:49.047272402Z: [INFO] ** (stop) :tcp_closed 2023-08-28T04:42:4…

How many connections are you opening? Azure has a limit on the amount of outgoing connections you can have. You need to check the limit for your instance size. You should also have a look around in the “Diagnose and solve problems” menu, it sometimes actually contains useful data. Personally the ne…

It’s just a new Phoenix app with the Repo configured to a SQL Server on Azure SQL. It’s running in a Docker container on Azure Web App or Web Service for Linux. There should be no users on it. The errors are logged less frequently if I set the pool_size to 1 instead of 10. Otherwise, the tcp_closed…

You are getting less errors because there are less database connections that are getting closed. I would still try to connect with the TDS driver directly, maybe even from your own machine to the Azure DB.

I’m able to connect to the Azure SQL database and run queries, but intermittently I get these errors. 2023-08-29T14:54:56.513337330Z ** (exit) an exception was raised: 2023-08-29T14:54:56.513342430Z ** (DBConnection.ConnectionError) bad return value: {:error, :closed} 2023-08-29T14:54:56.513348…

I was able to connect to the database with Tds on iex in the container and run queries. I get the DBConnection.ConnectionError from above, 1 out of 100 times.

Maybe try to execute some queries from a container, but using mssql-cli?

I changed the pool size to 1 and narrowed the TCP close errors to about every 90 to 100 seconds. I need to check if DBConnection is keeping the TCP connection alive or if Docker and Debian close connections in that time.

The application makes two TCP connections to the database on two different IP and ports. One connects on port 1433 and the other connects to a random port like 11061. I have the same Docker container running on my local and on Azure. On my local, traffic on both ports are constant every couple of …