Can I skip a corrupt archivelog in a restore?

Printable View

10-17-2003, 10:20 PM
eddieh

Can I skip a corrupt archivelog in a restore?

We are running an Oracle 8.1.7.4 EE instance on a server which lost its data RAID partition. I restored all the datafiles from an old backupset and upon applying the archivelogs (svrmgrl>>recover database until cancel using backup controlfile). I recieved the following:

ORA-00300: cannot open archived log 'nameoflogfile'
ORA-27047: unable to read the header block of file
OSD-04001: invalid logical block size (OS 47608)

Upon inspection we concluded that the archive log is corrupt and we have no other copies of it.

Is there a way I can skip this log file and apply the next 50 or so archivelogs I still have?

thanks in advance
10-17-2003, 10:25 PM
stmontgo

nope
10-17-2003, 10:35 PM
marist89

No way dude. You are limited to recover through your last valid archived redo log.
10-17-2003, 11:09 PM
eddieh

Not the answer I wanted to hear but thanks for the responses fellas.
10-18-2003, 12:45 AM
slimdave

Does anyone know of a way of validating the integrity of an archived log file, other than trying to restore it?
10-18-2003, 02:56 AM
adewri

Re: Can I skip a corrupt archivelog in a restore?

Quote:

Originally posted by eddieh
We are running an Oracle 8.1.7.4 EE instance on a server which lost its data RAID partition. I restored all the datafiles from an old backupset and upon applying the archivelogs (svrmgrl>>recover database until cancel using backup controlfile). I recieved the following:

ORA-00300: cannot open archived log 'nameoflogfile'
ORA-27047: unable to read the header block of file
OSD-04001: invalid logical block size (OS 47608)

Upon inspection we concluded that the archive log is corrupt and we have no other copies of it.

Is there a way I can skip this log file and apply the next 50 or so archivelogs I still have?

thanks in advance

Quote from metalink.

Quote:

Error: ORA 300
Text: illegal redo log block size specified - exceeds limit of
-------------------------------------------------------------------------------
Cause: The specified block size of the redo log file is greater than the
maximum block size for the operating system.
Action: Create the redo log on a device with a smaller block size.

By any chance was the new RAID filesystem created with different OS block size, as you have mentioned that you have lost the RAID???
10-18-2003, 03:19 AM
eddieh

We are using the default Win2k block size which has worked fine. The corrupt archivelog was actually a deleted file that I recovered using a disk recovery tool.
10-18-2003, 08:22 AM
marist89

Quote:

Originally posted by slimdave
Does anyone know of a way of validating the integrity of an archived log file, other than trying to restore it?

I know RMAN validates blocks on the way in for datafiles, but I'm not sure for redo entries.
10-19-2003, 02:39 AM
julian

Quote:

Originally posted by marist89
I know RMAN validates blocks on the way in for datafiles, but I'm not sure for redo entries.

RMAN will not backup a corrupted archive log file.

In Oracle9i, there can be up to 10 archived log destinations, so if an error occurs while reading one of them, Oracle will switch to another copy of that file.