tempdb error

Question

tempdb error

tcronin 95651

SSCertifiable

Points: 6735
More actions
February 18, 2016 at 6:35 am

#324426

I have a monster box has one OLAP instance and one SQL Instance(houses source data for OLAP). WHen processing one of the cubes i get
OLE DB error: OLE DB or ODBC error: The operating system returned error 23(Data error (cyclic redundancy check).) to SQL Server during a read at offset 0x000005709a0000 in file 'T:\tempdb\tempdb.mdf'
This is a 32 core Numa monster (new dell server) with SSD for the drives. I have 8 tempdb files (had 4 got same error). The tempdb disk is buried at 100%. This error show for data shows a corruption, ( i have 200 gig for tempdb) This is all SAN attached. ANy ideas?

Viewing 7 posts - 1 through 6 (of 6 total)

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

Gail Shaw SSC Guru Points: 1004504 More actions · Answer 1

Faulty SSD? Faulty switch/controller/cables (as applicable)?

Gail Shaw
Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability

We walk in the dark places no others will enter
We stand on the bridge and no one may pass

tcronin 95651 SSCertifiable Points: 6735 More actions · Answer 2

Hope to avoid going down that rabbit hole. The whole disk subsystem is SSD, was not formatted with 64K offset like I wanted, not to sure how much that makes a difference with SSD. Interesting that they managed to bury the drive with TEMPDB but my Idera snapshot showed me I had 40 gig of memory and CPU was at 20%. Almost like something got forced to use TEMPDB even though mem/cpu were available

Gail Shaw SSC Guru Points: 1004504 More actions · Answer 3

There's lots of things that will spill to TempDB, and even if there's available buffer pool memory they get written out (kept in memory, but written out). But if you're getting IO errors, that's an IO problem not one that's going to be solved by looking in SQL.

And did Ideara say 40GB of memory that nothing's using? If so, is SQL allowed to use it (via max server memory)

Gail Shaw
Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability

We walk in the dark places no others will enter
We stand on the bridge and no one may pass

tcronin 95651 SSCertifiable Points: 6735 More actions · Answer 4

max is set, I am not an SSAS expert and what it is doing to the resource instance, will start digging. Might try simple approach of having them spin up different drive and move TEMPDB. Also considering since this is physical of looking at the idea of the local tempdb drives (SSD) in the nodes (active/passive cluster). I have heard some good gains with this and you can get some fairly large disk for good prices. Better to be challenged then bored..........

tcronin 95651 SSCertifiable Points: 6735 More actions · Answer 5

Developer changed something in SSAS to persistent state and it fixed it

kevaburg SSCoach Points: 18131 More actions · Answer 6

I would be hesitant to say it that it is fixed. The error seemed to be quite obvious....

Apart from the fact you might not see the error at the Moment, what makes you think it is fixed?