Open DamionD opened 4 years ago
@DamionD It might be curious to setup a file watcher to ensure that there isn't something else scanning your file or utilizing it in some capacity that may cause the blocking.
In my setup I also have the DB on a CIFS and very occasionally I will get a write error with the same message when I have one process trying to write (typically me on command line) and other automated processes are accessing the DB of which I probably have 100s of calls a minute.
It could be the sleep is just spacing out your calls enough to avoid the other potential processes.
Can you replicate this on a test DB that is not used by anything else?
@DamionD It might be curious to setup a file watcher to ensure that there isn't something else scanning your file or utilizing it in some capacity that may cause the blocking.
In my setup I also have the DB on a CIFS and very occasionally I will get a write error with the same message when I have one process trying to write (typically me on command line) and other automated processes are accessing the DB of which I probably have 100s of calls a minute.
It could be the sleep is just spacing out your calls enough to avoid the other potential processes.
Can you replicate this on a test DB that is not used by anything else?
Hi @jkdba, this testing was performed on a brand new test database. The SMB server has Sophos anti-virus with on-demand scanning enabled; could this be relevant?
I have found the following issue when adding multiple entries to a KeePass database stored on an SMB share:
The process manages to add 1 to 4 of the entries before failing. Adding the entries individually works normally, as it does if I add a delay:
The minimum delay that appears to be needed is 11 milliseconds.
EDIT: The delay required seems to increase with the number of entries being processed (and possibly network latency).