Open sylvlecl opened 6 years ago
What do you mean by database cleanup?
Answering 2 years later :) : the database had been plainly deleted. For sure this particular situation will not happen in a managed production environment, but other situations may occur
Another issue observed in production environment, on Cassandra AFS implementation :
So it seems that cassandra tables are not consistent anymore, without being able to identify what has caused this.
Note that diagnostic and repair tools will be specific to an AFS implementation, so we may need to think of something very generic in the API.
A tool would be needed to check AppStorage data consistency, and to fix it in case of problems.
Such a problem is orphan nodes. In some circumstances, it may happen that some node has an invalid parent identifier (identifier of a node which has been deleted). At the moment, we don't have proper tools to detect and fix it.
That problem has happened in a real life situation where a service using the remote implementation of AFS has not been restarted while the database was cleaned up. That led to the service using an obsolete identifier for the root node (root id is kept on client side in the AppFileSystem class).