RE: db corruption

Something I've noticed lately, in two different databases now:

Oracle 8.1.7.4 and 9.2.0.6, Solaris Sparc, raw volumes, served up from an EMC DMX, via VxVM.

Rman reports corruption. It says the block is all zeroes. Alert log shows that it tried to read the block, tried to read the mirror, same error. Actual block dump indicates the block is in fact all zeroes. The block is *always* the last block in the datafile, and doesn't contain any data. So, I set corruption count to 1, backup the datafile, restore it, recover it, and RMan formats/fixes the block on restore. I've had one database where this occurred once. I've had another where this happened to 60-70 datafiles. Exact same type of corruption, always the last block in the datafile.

Anyone else ever seen this?

AdvThanksAnce,

-Mark

--
Mark J. Bobak
Senior Oracle Architect
ProQuest Information & Learning

Ours is the age that is proud of machines that can think and suspicious
of men who try to.  --H. Mumford Jones, 1892-1980


-----Original Message-----
From: oracle-l-bounce_at_freelists.org
[mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Kevin Closson
Sent: Tuesday, August 15, 2006 12:45 PM
To: ORACLE-L
Subject: RE: db corruption

 >>>


>>>I recently had rman complaining about corruption in archivelogs. I 

>>>dumped the archivelogs, no corruption. The sysadmin checked the 

>>>disks, they were fine. He sent the T3 logs to sun. They said there 

>>>were a couple of really soft errors on the disks and if rman was 

>>>super-sensitive it might show. They changed the disks and now all is 

>>>good.


This is weird. RMAN cannot be "super-sensitve" to anything ("really
soft"--whatever that is--or otherwise). Either its I/O calls fail or
they don't. I don't doubt it when you say "all is now good", but the
relief likely does not come from swapping out drives that are suffering
soft errors.
 
Golly is support getting spammy these days or what?
--
http://www.freelists.org/webpage/oracle-l


--
http://www.freelists.org/webpage/oracle-l