May 23, 2005 at 10:44 am
Hi,
I attempted to restore a database today and it restored very quickly compared to usual, and the normal completion message appeared. I noticed that the database was showing as loading (which never completes) and checked the error log, which had the following message:
I/O error on backup or restore restart-checkpoint file
'e:\resource\MSSQL\backup\R32Train.ckp'.
Operating system error 3(error not found).
The statement is proceeding but is non-restartable.
The server is definitely not full because we cleaned 5GB last week and no new databases have been created since then.
I would really appreciate any help on this - I need to ensure that it does not happen again
Thanks very much,
Roly
May 23, 2005 at 12:07 pm
Forgive I'm stating the obvious, but OS error 3 is a 'Path not found' error. Have you verified that the file actually exists in the folder where SQL Server is looking for it?
Other than that, are you running Anti-Virus or 3rd party backup tools on this server? If yes, make sure that these processes aren't interfering with your restore. With A/V software it's normally a best practice to exclude your *.mdf and *.ldf files from scans as they can cause numerous problems. Third party backup and restore tools can keep files open exclusively when SQL Server attempts to access them and cause similar error messages. Hope this helps.
My hovercraft is full of eels.
May 23, 2005 at 12:56 pm
Hi,
The error message says "error not found" - I noticed the "path not found error" on the internet, but I had already checked that the file exists through windows explorer. I have just checked again and the backup is there.
The problem is not being experienced on the server itself, just when I tried through the network. I'm not part of the IT department and have experienced no problems up until now.
Thanks for your help,
Roly
May 23, 2005 at 1:57 pm
Interesting. 'Error not found' isn't much help, is it?
Can you contact a network admin and have them verify the restore status by running Enterprise Manager directly from the console? Or is it possible for you to open a terminal server session to the server to verify the status from there? I'm assuming that you've restarted the system that you originally started the backup from and have refreshed your connections.
I did some searching on this specific error message as well, and most people that seemed to get it had a problem with Veritas backup software. Is it possible that this was installed on your server by the IT department without you knowing about it? The fact that you state "the database restored more quickly than usual" makes me suspicious that something has changed or been done on the server end.
My hovercraft is full of eels.
May 24, 2005 at 8:06 am
Have you checked to see if the backup file actually has data in it? If a backup fails or is interrupted then the backup file could still exist with no data or short data.
-n
May 24, 2005 at 9:46 am
Is e: a drive on your system or is it a drive on the server. You can only backup and restore to drives that are considered 'local' by the SQL server.
July 18, 2005 at 12:14 pm
I had a similar problem awhile ago. The only way that I could find to overcome it was to disable Named Pipes, and only use TCP/IP.
Viewing 7 posts - 1 through 6 (of 6 total)
You must be logged in to reply to this topic. Login to reply