home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.vms
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!news.acns.nwu.edu!casbah.acns.nwu.edu!jweiss
- From: jweiss@casbah.acns.nwu.edu (Jerry Weiss)
- Subject: Re: Long standing apparent Backup or XQP problem
- Message-ID: <1993Jan6.182749.27733@news.acns.nwu.edu>
- Sender: usenet@news.acns.nwu.edu (Usenet on news.acns)
- Nntp-Posting-Host: unseen1.acns.nwu.edu
- Organization: Northwestern University, Evanston Illinois.
- References: <9301052330.AA21443@relay1.UU.NET>
- Date: Wed, 6 Jan 1993 18:27:49 GMT
- Lines: 58
-
- In article <9301052330.AA21443@relay1.UU.NET> gleeve@galaxy.dnet writes:
- [misc deleted]
- >
- >What I did:
- > I ran Backup on a VAX 6660 to create savesets, on the same disk,
- > where the disk had plenty of space (well, it had plenty before
- > I started :-) ). The commands were of the form
- >
- >$ backup/block=4096/trunc [.a*...] [somewhere]a.bck/save
- >
- >which normally creates a saveset of all the relevant directories.
- >
- > 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.
- >
-
-
- > The point of all this is cautionary: at least SOME of the
- >time use BACKUP/COMPARE or BACK/LIST on your savesets to ensure
- >this is not happening to you.
- > If anyone can shed light on the problem beyond this caution,
- >please feel welcome. If it happens again I will gather more information
- >on the account. Still, the problem didn't occur, with the same
- >account, where backup was running on a uniprocessor.
-
- This may not be relevant, but I've seen a few odd things in my backups as
- well. I started to run them with a /verify pass to give me more confidence
- in them. Most of my backups are to mag tape, 8mm emulation of TMSCP,
- usually run with /NOCRC/GROUP=0. I understand the downside of the last two
- switches, but its a decision I've made for myself and the benefits/hazards
- are probably best handled in a seperate thread. What gets reported as a
- hard or soft error for a tape may not have anything to do with what Glenn
- mentioned above.
-
- 1) SOFTERRS, recoverable media errors. Not logged in errorlog or show
- error. Maybe this is just normal.
-
- 2) Verify errors on files that are closed and never accessed. Verify
- errors disappear when I run with /CRC. The incident of errors is
- low ~1/800,000 blocks. I was beginning to suspect that data going
- between controller and drive was getting corrupted. I would have
- thought the /CRC (with /Group=0) would have caused the appearance of
- some read errors in that case.
-
- 3) Interaction between MSA0 (TU80) and MUA0:. Mounts of MSA0 appear to
- be blocked during certain MUA0 operations (which can take considerable time).
-
- This is on a 11/750 VMS 5.4-3 with CSCPAT_010116 patches applied.
-
-
-
- --
- Jerry S. Weiss "If you can't stand the heat, stay out of the antimatter!"
- j-weiss@nwu.edu Dept. Medicine, Northwestern Univ. Medical School
-