App sometimes unbale to fetch data

[upgrader-dev]

Hello,

I developped a simple web app. Its goal is to allow people in a party to add songs to the queue via a link or a qrcode.

The app works flawlessly when I test it alone deployed or not or when I run tests.

But when my app is used by real users I often get an error saying that the app is unable to find a data in the db while this data exists in the db.

{"message":"Uncaught PHP Exception Error: "Call to a member function getId() on null"

The object is null while it should not because data is indeed present in the db. I checked that by running queries in the db service of the container.

And a few seconds later The app is able to fetch the data.

After a few un successful tries, I have to completely compose down and up to retrieve my app.

Initially I thought that I had reached the ceiling of the requests my VPS could handle so I upgraded my plan by two steps from 5€ a month to 15€ a month.

This happens usually when I reach at least 20 users in the party.

It seems that the app has difficulty to communicate with the mysql service.

Maybe I should change to postgres or something else.

I have already ran docker system prune and have no unused images on the vPS.
Any idea?

[upgrader-dev]

To begin I will look at the queries sent to the database.
Under the dev environment I have nice debugging tools but obviously not under prod env.

Anyone know how I can access to the queries made to the database service?

[7-zete-7]

Hi @upgrader-dev!

It seems to be more related to the infrastructure or environment you are using.

I can only assume, given that the database service is part of this project, that the VPS you are using periodically scales your project and, at the time of initialization, there are moments when the database is already available, the project is already responding, but the database is not yet initialized with data.

In any case, it looks like the issue is related to this template.

[upgrader-dev]

Hi @upgrader-dev!

It seems to be more related to the infrastructure or environment you are using.

I can only assume, given that the database service is part of this project, that the VPS you are using periodically scales your project and, at the time of initialization, there are moments when the database is already available, the project is already responding, but the database is not yet initialized with data.

In any case, it looks like the issue is related to this template.

Hi 7-zete-7

Yeah initially I was thinking about that.
But I have learnt to not always think that the issue is coming from the anything else than my work😊.
And I have already tried to upgrade my VPS plan without success.

I will contact the support of my provider.

Apart from that if I want to validate your thoughts I should be able to find some logs confirming that my database service is unreachable or something like that, but I didn't. Do you know where I could find such logs?

Another solution would be to write my code so the queries wait for the database to be reachable. But that's definitely not my first approach as it will decrease the overall performance.

[7-zete-7]

<...> Do you know where I could find such logs?

Logs for the database service can be obtained using the docker compose logs database command (see https://docs.docker.com/reference/cli/docker/compose/logs/). This command can return logs for the entire lifetime of the container. If the cause was restarts of the database service, there should be information about the restarts in the logs.

---

Another solution would be to write my code so the queries wait for the database to be reachable. But that's definitely not my first approach as it will decrease the overall performance.

There should be no need for this:

• Containers of official database images should work stably and not restart accidentally. If a container did restart, it will be visible in the logs of their container.

• When starting the php service, there is already a check for the availability of the database service.
  

  symfony-docker/frankenphp/docker-entrypoint.sh

  Lines 34 to 53 in 1b01814

  	echo 'Waiting for database to be ready...'
  	ATTEMPTS_LEFT_TO_REACH_DATABASE=60
  	until [ $ATTEMPTS_LEFT_TO_REACH_DATABASE -eq 0 ] || DATABASE_ERROR=$(php bin/console dbal:run-sql -q "SELECT 1" 2>&1); do
  	if [ $? -eq 255 ]; then
  	# If the Doctrine command exits with 255, an unrecoverable error occurred
  	ATTEMPTS_LEFT_TO_REACH_DATABASE=0
  	break
  	fi
  	sleep 1
  	ATTEMPTS_LEFT_TO_REACH_DATABASE=$((ATTEMPTS_LEFT_TO_REACH_DATABASE - 1))
  	echo "Still waiting for database to be ready... Or maybe the database is not reachable. $ATTEMPTS_LEFT_TO_REACH_DATABASE attempts left."
  	done
  	if [ $ATTEMPTS_LEFT_TO_REACH_DATABASE -eq 0 ]; then
  	echo 'The database is not up or not reachable:'
  	echo "$DATABASE_ERROR"
  	exit 1
  	else
  	echo 'The database is now ready and reachable'
  	fi

---

Can you also make sure that you have the container's data directory mounted to a volume or filesystem?
For the mysql image, this should be the /var/lib/mysql directory (see the "Where to Store Data" section on https://hub.docker.com/_/mysql).

[upgrader-dev]

The bug occured again this afternoon. I got these lines in the app image
{"message":"Uncaught PHP Exception Doctrine\DBAL\Exception\ConnectionLost: "An exception occurred while executing a query: SQLSTATE[HY000]: General error: 2006 MySQL server has gone away" at ExceptionConverter.php line 105","context":{"exception":{"class":"Doctrine\DBAL\Exception\ConnectionLost","message":"An exception occurred while executing a query: SQLSTATE[HY000]: General error: 2006 MySQL server has gone away","code":2006,"file":"/app/vendor/doctrine/dbal/src/Driver/API/MySQL/ExceptionConverter.php:105","previous":{"class":"Doctrine\DBAL\Driver\PDO\Exception","message":"SQLSTATE[HY000]: General error: 2006 MySQL server has gone away","code":2006,"file":"/app/vendor/doctrine/dbal/src/Driver/PDO/Exception.php:24","previous":{"class":"PDOException","message":"SQLSTATE[HY000]: General error: 2006 MySQL server has gone away","code":0,"file":"/app/vendor/doctrine/dbal/src/Driver/PDO/Statement.php:130"}}}},"level":500,"level_name":"CRITICAL","channel":"request","datetime":"2025-04-01T13:19:43.606908+00:00","extra":{}}

Unfortunately docker logs returns nothing for the moment. I have to manually enable logging of the mysql image.

[7-zete-7]

@upgrader-dev, it seems to be related to php/frankenphp#290.

[upgrader-dev]

Ok seems like I have to upgrade to Symfony 7.1 (6.4 for the moment). I'll give it a try