latch timeout
-
Hi All,
I am getting a latch (Type 4) timeout when accessing a certain page.
CheckDB returns no errors and after truncating the table (Even dropping it) I still get the error.
Trying to shrink the DB causes the error as well as all the backups.
I got a log backup that runs once an hour and it's failing.
Any ideas what would be wrong? Corrupt page?
A CHKDSK returned no errors either.
Cheers,CrispinI can't die, there are too many people who still have to meet me!It's not a bug, SQL just misunderstood me!
-
Hi
If you have a corrupt page, u can find something in the sql server logs. Check EM, management, sql server logs for occurance of some kind of error concerning datapages.
hth
JP
-
Forgot to mention that...
There was an assertion error and SQL died. An insert was taking place.
One of the guys restarted the service and thought nothing of it.
My problem is not so much what happened but how do I fix it now?
Cheers,CrispinI can't die, there are too many people who still have to meet me!It's not a bug, SQL just misunderstood me!
-
get the database into 'single user mode' and runn a couple of DBCC commands. My guess is that DBCC CHECKDB will be clean. I'd try the following:
DBCC CHECKALLOC ( 'database_name' , REPAIR_REBUILD )
WITH ALL_ERRORMSGS
Before anything I'd talk to the OS folks and have them run diagnostics first on memory and disk. You should probably look at the System and Application event logs first before contacting them.
RegardsRudy KomacsarSenior Database Administrator"Ave Caesar! - Morituri te salutamus."
-
I did run a checkAlloc but with out the switch. It returned no errors. Will run it this evening.
There's nothing out of the ordinary in the event log other than SQL complaining about the latch timeout.
Unfortunatly, I wear the "OS guy" hat as well
Will see what the outcome of the DBCC is...
Cheers,CrispinI can't die, there are too many people who still have to meet me!It's not a bug, SQL just misunderstood me!
-
A simple reboot.....
I found a lot of people on google groups talking about this. A couple of them said it was a conflict between the cache on the controller and disk.
Problem occured while doing a check point. Any normal reading of the page was ok.
Anyway, rebooted, changed from simple to full and backed up fine.
Chalk it up to another "Hope it does not happen again" things.
Cheers,CrispinI can't die, there are too many people who still have to meet me!It's not a bug, SQL just misunderstood me!
Viewing 6 posts - 1 through 5 (of 5 total)
You must be logged in to reply to this topic. Login to reply