Status 3. File Not Open.

[JoshBreitbach]

I have been having a lot of issues lately related to continuous operations backup routines. What happens is our data files (approx 450) are put in to continuous operation mode at 6PM by using the butil -startbu command and a listfile. I have never had any error messages show up in the log for this command. Our backups are performed between 6 and 8-9 PM and we never encounter errors at this point.

We have chosen 10 PM as a safe time to take the DB out of BU mode. So at 10 PM, we perform a butil -endbu with a listfile. This is where the problem occurrs. I sometimes (once a week) get an error in my log file. The error is as follows:

BUTIL-14: The file that caused the error is H:\CPC_PARADIGM\BTRV\INV\FISCALWK.BTR,H:\CPC_PARADIGM\BTRV\INV\CITYDEST.BTR,H:\CPC_PARADIGM\BTRV\INV\REMITTO.BTR,H:\CPC_PARADIGM\BTRV\INV\INVCUST.BTR,H:\CPC_PARADIGM\BTRV\INV\INVOICE.BTR,H:\CPC_PARADIGM\BTR.
BUTIL-100: MicroKernel error = 3. The file is not open.

It appears that on the first line, it was listing all the files in the listfile but ran out of space. The second line is the actual error message. The affected file is usually different each time. Last time it was Data0075.btr (approx 100KB!).

Then any attempt to run any application that uses that file results in the Status 3 File not open error. This includes trying to do a butil -stat. The only way we have found to recover from this is to shut down the services and start them back up. This is a terrible solution because our database is in constant use and shutting down services requires everyone to drop out of the software etc.

Is there any way that anyone knows to prevent this error? Or at least a way to recover gracefully without shutting down services?

Pervasive 7 SP 6 running on Windows NT 4 with latest updates.

Thanks,
Josh Breitbach
jpbreitb@rockwellcollins.com

 

[TomKane]

Are you using netware? Are you doing this startbu and endbu purely for your backups? You should check the pervasive monitor to see if there are any configuration settings that are reaching their peaks - I think this gets reset after you start and stop the services. Maybe the thing to do is allocate time to do the backup make everyone get out of the system and do the backup and then allow the users back in again.