We had a hard drive failure in one of our RAID arrays. We've resolved the
hardware issue; however, DBCC CHECKDB is showing errors in at least some of
our databases. The way we structure our databases is with two data files,
the second containing all of our table indexes. The RAID array that failed
contained these "index-only" data files.
Since DBCC CHECKDB reported that the minimum repair option would be
'REPAIR_ALLOW_DATA_LOSS', I thought I would be restoring each database from
our last full plus differential backups. It's approximately 45 databases,
and I was hoping to avoid this.
Since CHECKDB seems to be reporting an error on only a single table in each
database AND the hardware failure was on the array that contains only
indexes, I thought simply reindexing the table reporting an error might be
an option. So I took one of the databases showing an error and reindexed
the offending table with DBCC DBREINDEX. It completed successfully, so I
reran the DBCC CHECKDB and it reported no errors.
I realize restoring from backups might be the safest route to go, but if I
only reindex a table and the dbcc checkdb comes up clean, am I at risk for
any data loss?
I appreciate your insight on this situation.
Thanks
-joeIf the error was in the index then rebuilding the index will in most cases
fix the problem. If you do a DBCC DBREINDEX you are rebuilding the index
from scratch. So as long as there are no hardware issues to prevent a clean
rebuild you should be OK. The problem comes into play more when it is data
corruption than indexes.
Andrew J. Kelly SQL MVP
"Joe Ross" <joeross7@.hotmail.com> wrote in message
news:O25cjrE7FHA.744@.TK2MSFTNGP10.phx.gbl...
> We had a hard drive failure in one of our RAID arrays. We've resolved the
> hardware issue; however, DBCC CHECKDB is showing errors in at least some
> of our databases. The way we structure our databases is with two data
> files, the second containing all of our table indexes. The RAID array
> that failed contained these "index-only" data files.
> Since DBCC CHECKDB reported that the minimum repair option would be
> 'REPAIR_ALLOW_DATA_LOSS', I thought I would be restoring each database
> from our last full plus differential backups. It's approximately 45
> databases, and I was hoping to avoid this.
> Since CHECKDB seems to be reporting an error on only a single table in
> each database AND the hardware failure was on the array that contains only
> indexes, I thought simply reindexing the table reporting an error might be
> an option. So I took one of the databases showing an error and reindexed
> the offending table with DBCC DBREINDEX. It completed successfully, so I
> reran the DBCC CHECKDB and it reported no errors.
> I realize restoring from backups might be the safest route to go, but if I
> only reindex a table and the dbcc checkdb comes up clean, am I at risk for
> any data loss?
> I appreciate your insight on this situation.
> Thanks
> -joe
>|||Thank you. This is how I understood it, but it's reassuring to get
confirmation from an expert.
Much appreciated,
-joe
"Andrew J. Kelly" <sqlmvpnooospam@.shadhawk.com> wrote in message
news:ej%23s64F7FHA.2484@.TK2MSFTNGP10.phx.gbl...
> If the error was in the index then rebuilding the index will in most cases
> fix the problem. If you do a DBCC DBREINDEX you are rebuilding the index
> from scratch. So as long as there are no hardware issues to prevent a
> clean rebuild you should be OK. The problem comes into play more when it
> is data corruption than indexes.
> --
> Andrew J. Kelly SQL MVP
>
> "Joe Ross" <joeross7@.hotmail.com> wrote in message
> news:O25cjrE7FHA.744@.TK2MSFTNGP10.phx.gbl...
>|||Andrew, There's another SQL 2005 launch event December 19th in Waltham if
you'd care to post it on NESQL.
Please email me if you'd like more details.
--
burt_king@.yahoo.com
"Joe Ross" wrote:
> Thank you. This is how I understood it, but it's reassuring to get
> confirmation from an expert.
> Much appreciated,
> -joe
> "Andrew J. Kelly" <sqlmvpnooospam@.shadhawk.com> wrote in message
> news:ej%23s64F7FHA.2484@.TK2MSFTNGP10.phx.gbl...
>
>
Friday, February 17, 2012
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment