From:	MERC::"uunet!CRVAX.SRI.COM!RELAY-INFO-VAX"  7-JAN-1993 15:14:33.33
To:	info-vax@kl.sri.com
CC:	
Subj:	Re: Long standing apparent Backup or XQP problem

In article <9301052330.AA21443@relay1.UU.NET>, gleeve@galaxy.dnet writes:
> When making up masters for the VMS/L&T sig tapes at the recent
> DECUS symposium, I encountered a problem making a Backup
...
>   Doing a BACKUP/LIST of [somewhere]a.bck/save
> 
> revealed errors recovered by XOR groups, and errors of lost blocks
> in files within the saveset, EVEN THOUGH there were NO errors reported
> by Backup during creation of the savesets, nor any hardware errors
> (via show error) reported on the device.
> 

Bingo! I had this problem backing up from a uVaxII to a 4500 via LAVC. The
problem was consistant, repeatable ( modulo the number of error ), and *VERY*
dependant on load. IE, Decterm in and no errors, use decterm to 'set host',
ave of 4-6 errors per RA-81, OPA0 6-10 errors...

My suspicion is a race condition in the SCS/MSCP/QIO code. This has been an
open problem report since Aug 92. 5.5 and 5.5-1 BTW.

I tryed /analing the savesets, but bloody backup hides the error!! Any one have
a hacked version that will dump the blockes in error they can send me?

~Paul