| Author |
Topic |
|
443dds
Starting Member
7 Posts |
 Posted - 2005-07-19 : 23:26:37
|
| We have multiple databases on our SQL box, with each one of them on a different volume on a MSA1000 (SAN). Our main database was getting too large to reindex on its allocated volume, so last week we added an Adaptec SANbloc and dedicated this hardware for this database. We just plugged it into a free port on our SAN switch, and away we went.We had a couple of application errors, so we ran DBCC CHECKDB to find the log below.I have no idea if we have a hardware issue with the new controller, if write cache is hosing things, or if we have just bad luck.Any ideas? Should we move this DB back to the old hardware? Will we be forced to restore, or will the DB be repairable? (We ran a repair_with_data_loss on PS_ITEM_DST [39 million rows) but it didn't fix the errors [it is possible they are new errors though.])Thanks.(The log doesn't all fit in this window, so I'll copy and paste into a couple different posts.) |
|
|
443dds
Starting Member
7 Posts |
Posted - 2005-07-19 : 23:30:24
|
| Really, the log is pretty big to even copy-paste. (3mb in plain text)It is available here in txt form:http://chogg.wwdd.net/sqllog.txt |
 |
|
|
franco
Constraint Violating Yak Guru
255 Posts |
Posted - 2005-07-20 : 03:57:13
|
| First of all I would check for errors at the Hardware level and at the Windows event log.No system errors reported there?Franco |
|
|
|
443dds
Starting Member
7 Posts |
Posted - 2005-07-20 : 04:25:15
|
quote:
Originally posted by francoNo system errors reported there?Franco
Nope. None in app or system log having to do with hardware errors. Goes back for a few days, though possibly not from the start of the issue. |
|
|
|
franco
Constraint Violating Yak Guru
255 Posts |
|
|
443dds
Starting Member
7 Posts |
Posted - 2005-07-20 : 05:34:12
|
| We are using Win2000 Advanced Server SP4 with SQL Enterprise SP4. |
|
|
|
443dds
Starting Member
7 Posts |
Posted - 2005-07-20 : 05:59:16
|
| When performing a dbcc checktable ('PS_ITEM_DST',REPAIR_ALLOW_DATA_LOSS)I get this:http://chogg.wwdd.net/allowloss.txt |
|
|
|
franco
Constraint Violating Yak Guru
255 Posts |
Posted - 2005-07-20 : 08:44:42
|
| Have you check your HDD's?Franco |
|
|
|
443dds
Starting Member
7 Posts |
Posted - 2005-07-20 : 14:14:07
|
| We are using a SAN enclosure, and the enclosure has logged zero errors. |
|
|
|
franco
Constraint Violating Yak Guru
255 Posts |
Posted - 2005-07-21 : 02:23:03
|
| I had a SAN disk in a dirty state but for the Storage Manager everything was ok, no errors logged.I had to run chkdsk on a clustered shared disk in order to solve my problem.In that case events log where notified only on the Windows event log:The file system structure on the disk is corrupt and unusable. Please run the chkdsk utility on the volume R:Maybe your situation is different.Franco |
|
|
|
paulrandal
Yak with Vast SQL Skills
899 Posts |
Posted - 2005-07-21 : 11:12:15
|
quote:
Originally posted by 443dds
When performing a dbcc checktable ('PS_ITEM_DST',REPAIR_ALLOW_DATA_LOSS)I get this:http://chogg.wwdd.net/allowloss.txt
Just looked at these - Franco's right - your disk(s) are definitely bad. All these pages have garbage overwriting the page headers. Repair has deallocated all of these pages so you've lost all the data that was on them.Paul RandalDev Lead, Microsoft SQL Server Storage Engine |
|
|
|
443dds
Starting Member
7 Posts |
Posted - 2005-07-21 : 12:13:01
|
| Since I am using a RAID controller to manage the disks, is it the controller's fault, or can a specific disk in the array be a problem as well? |
|
|
|
franco
Constraint Violating Yak Guru
255 Posts |
Posted - 2005-07-22 : 03:35:35
|
quote:
Originally posted by 443dds
We are using a SAN enclosure, and the enclosure has logged zero errors.
If your enclousure reports zero errrors, as you said, I would check at the HDD level.Franco |
|
|
|
|
Consistency Errors