Post reply

RESTORE DATABASE FAILS with Write on "???" failed: 1117(The request could not be performed because of an I/O device error.)

  • July 3, 2012 at 10:37 am

    #392650

    I am currently fighting an issue in which I'm looking for confirmation that this could be related to a bad backup or not.

    I have attempting to restore two seperate full backs for the same database.

    What I am seeing is the following error during the process in the sql logs:

    SQL Server has encountered 1 occurrence(s) of I/O requests taking longer than 15 seconds to complete on file [I:\database_name.mdf] in database [<Database name is not available>] (5). The OS file handle is 0x00000000000007EC. The offset of the latest long I/O is: 0x00001adda30000

    Then the process a little later fails with:

    Msg 3202, Level 16, State 1, Line 1

    Write on "???" failed: 1117(The request could not be performed because of an I/O device error.)

    Msg 3013, Level 16, State 1, Line 1

    RESTORE DATABASE is terminating abnormally.

    The system is virtual. One thing we have attempted to do is to remove the vdmk and recreated it in the same datastore and lun on the san within a share raid group.

    Backups are being performed by the standard backup maintenance plan. Compression is enabled and verfiy backup integrity is enabled.

    The server is running OS Windows Server 2008 R2 and SQL Server 2008 R2 SP1.

    I believe this is being cause due to an I/O issue with the disk subsystem. However, could this be cause due to a bad backup?

  • July 3, 2012 at 10:46 am

    #1508336

    Don't think so. The first error is an I/O subsystem error, surfaced when the I/O system doesn't respond correctly.

    The backup error would surface differently as a failure to read from offset xxxx in the backup file.

  • July 3, 2012 at 10:57 am

    #1508346

    Thanks Steve. Just wanted to make sure we were spinning the cycles in the right location.

  • July 3, 2012 at 4:55 pm

    #1508538

    I just wanted to post a followup on this.

    We ran three test to the backup

    1. Restored to a Physical Server

    2. Restored to a another virtual server using nas instead of San

    3. Restored to the same virtual server after creating a new lun, datastore and virtual disk.

    The compress backup we were restoring was 33 Gb to expand to a 140 gb mdf and 17 gb ldf

    1. Took about 40 minutes to complete without any timeout messages

    2. Took 4 Hours 30 minutes without any timeout messages

    3. Took about 55 minutes with the same timeout message at about 40% complete

    So with this I think it was safe to confirm that the issue was related to the drive I was restoring to and not the backup.

    Again, Steve thanks for confirming where to focus the efforts.

  • July 3, 2012 at 5:00 pm

    #1508539

    You are welcome, and glad things are working for you.

    Steve

  • July 4, 2012 at 5:28 am

    #1508721

    Also has more or less similar problem where the issue was not even the backup but the log file.

    We couldn';t backup the log file and it was increasing at least by 25GB a day.

    Couldn't do a log back up but a full DB backup would backup succesful.

    When doing a log backup I'd get error below:-

    Msg 3203, Level 16, State 1, Line 1

    Read on "L:\Program Files\Microsoft SQL Server\MSSQL10_50.MSSQLSERVER\MSSQL\Data\FILENAME_log.ldf" failed: 1117(failed to retrieve

    text for this error. Reason: 15105) Msg 3013, Level 16, State 1, Line 1BACKUP LOG is terminating abnormally.

    This turned out to be something wrong with the disk where the log file was placed.

    What you don't know won't hurt you but what you know will make you plan to know better

  • July 10, 2012 at 2:31 am

    This was removed by the editor as SPAM

  • July 10, 2012 at 2:39 am

    This was removed by the editor as SPAM

Viewing 8 posts - 1 through 7 (of 7 total)

You must be logged in to reply to this topic. Login to reply