November 25, 2011 at 4:52 am
Thanks everyone for responding.
The problem came back (different DB's trans. log this time).
As for the questions, no there are no other I/O related errors on the SQL server or the server itself. And yes, the backup is done using maintenance plan.
I'm running various HW diagnostics at the moment, all came clean so far :/
One more question - can there be a log corruption that doesn't originate from HW problem (such as bad sectors mentioned here) like a malformed query in an application using the DB? Or does this always revolve around HW problems?
November 25, 2011 at 7:20 am
Rambler (11/25/2011)
Thanks everyone for responding.The problem came back (different DB's trans. log this time).
As for the questions, no there are no other I/O related errors on the SQL server or the server itself. And yes, the backup is done using maintenance plan.
I'm running various HW diagnostics at the moment, all came clean so far :/
One more question - can there be a log corruption that doesn't originate from HW problem (such as bad sectors mentioned here) like a malformed query in an application using the DB? Or does this always revolve around HW problems?
Could be, I've once hit & found a rollback bug that corrupted the DB.
That being said it's probably a 1 billion to 1 chance of being the culprit.
It's "always" the HW, somewhere between sql server and itself again.
December 2, 2011 at 3:41 am
Just an update, even Dell (the server manufacturer) didn't find any hardware problem.
Looks like, I'll have to open a case with Microsoft, not sure they'll be of any help though.
January 13, 2012 at 1:16 pm
Curious to know what ever came of this. Hopefully Rambler can give us an update.
January 13, 2012 at 6:41 pm
Unfortunately nothing fruitful. The problem got dissolved in company-client politics since there was no conclusive proof. Actually the transaction log corruption happened again after couple weeks of everything going well, but I got "overruled" by superiors that it's not a HW problem and if it is, it'll be solved by switching the RAID controller for new (and faster with more cache) and splitting the DB a TLs to separate disks (which should've been done from the start anyway).
If I have anything worth updating, I'll post it here.
January 13, 2012 at 6:50 pm
Rambler (1/13/2012)
Unfortunately nothing fruitful. The problem got dissolved in company-client politics since there was no conclusive proof. Actually the transaction log corruption happened again after couple weeks of everything going well, but I got "overruled" by superiors that it's not a HW problem and if it is, it'll be solved by switching the RAID controller for new (and faster with more cache) and splitting the DB a TLs to separate disks (which should've been done from the start anyway).If I have anything worth updating, I'll post it here.
That'll work about as good as putting a smiley face on your fuel warning light and NOT stopping for gaz.
I hope they hit the correct solution even without investigating!
March 28, 2012 at 3:54 am
Yes they are totally two different things however they share common thing which is data processing. the Log is responsible for transactions that are not yet committed on the Database, so if u have data that's been processed it will be sitting here until all transactions are completed.
Just go read about the ACID properties of a Transaction to get further clarity.
Viewing 7 posts - 16 through 21 (of 21 total)
You must be logged in to reply to this topic. Login to reply