You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We are using Firebird databases for automatic integration tests. We create all our test databases using a script. For each test a copy is made of one of the test databases. The test program only connects to the copy. When the test is complete, the copy is deleted. This way we guarantee that the different test programs do not interfere with each other. We also run a lot of the tests in parallel, all against the same server, but each against a database of their own.
The problem is that sometimes the server hangs and the only way to get it working again is to restart it.
Sadly the problem is not really reproducible. It can work fine for weeks, then hang several times in one day.
My *guess* to why this happens is that when we clean up after a test, then sometimes the test program has not finished closing the database connection. This can happen if a test is killed, which happens automatically when it takes too long. It might be that the server has the database open when we delete the database file and also the directory where it resides.
The text was updated successfully, but these errors were encountered:
Submitted by: Håkan Johansson (hakjo)
We are using Firebird databases for automatic integration tests. We create all our test databases using a script. For each test a copy is made of one of the test databases. The test program only connects to the copy. When the test is complete, the copy is deleted. This way we guarantee that the different test programs do not interfere with each other. We also run a lot of the tests in parallel, all against the same server, but each against a database of their own.
The problem is that sometimes the server hangs and the only way to get it working again is to restart it.
Sadly the problem is not really reproducible. It can work fine for weeks, then hang several times in one day.
My *guess* to why this happens is that when we clean up after a test, then sometimes the test program has not finished closing the database connection. This can happen if a test is killed, which happens automatically when it takes too long. It might be that the server has the database open when we delete the database file and also the directory where it resides.
The text was updated successfully, but these errors were encountered: