Daily I check for tapes that are readonly or unavailable...
I'll then do a "q act begind=-1 s=<volser>" which might or might not
provide an ANRxxxx[E,S,D] message number...
but if I turn around and do another "q act begind=-1 begint=<close to the
last time the tape was mounted> s=<rmt# the tape was last mounted on>"
I can often find out more about what went wrong...
example from recently...
(Embedded image moved to file: pic14589.jpg)
the anr8469e is where the tape failed to dismount... then just a few days
later the anr8359e is where it was having media faults...
I didn't like the combination of these errors this close together so I
pulled the tape. (oh, the volser was in column1 but its name was removed
to protect the innocent ;-)
Dwight
Joni Moyer
<joni.moyer@HIGHM To: ADSM-L AT VM.MARIST
DOT EDU
ARK.COM> cc:
Sent by: "ADSM: Subject: Re: When to declare a
tape dead
Dist Stor
Manager"
<[email protected]
.EDU>
04/29/2003 10:02
AM
Please respond to
"ADSM: Dist Stor
Manager"
How do you know what ANR message numbers to look for? Thanks!
Joni Moyer
Systems Programmer
joni.moyer AT highmark DOT com
(717)975-8338
Dwight Cook
<cookde AT US.IBM DOT CO To:
ADSM-L AT VM.MARIST DOT EDU
M> cc:
Sent by: "ADSM: Subject: Re: When to
declare a tape dead
Dist Stor
Manager"
<[email protected]
.EDU>
04/29/2003 10:50
AM
Please respond to
"ADSM: Dist Stor
Manager"
I keep a spread sheet with the volsers in column A
then each time a tape encounters an error, I log the date & the ANRxxxx_ in
a following column.
remember, often tape problems are only/actually reflecting tape DRIVE
problems...
Anyway, with time, as I complete this spread sheet...problem tapes stand
out REAL FAST !
Key is to track your problems and what sort of problems they are...
Dwight
"Lambelet,Rene,VE
VEY,GL-CSC" To:
ADSM-L AT VM.MARIST DOT EDU
<Rene.Lambelet@NE cc:
STLE.COM> Subject: Re: When to
declare a tape dead
Sent by: "ADSM:
Dist Stor
Manager"
<[email protected]
.EDU>
04/29/2003 04:37
AM
Please respond to
"ADSM: Dist Stor
Manager"
HI, we do the same except we move leave the tape in scratch pool after move
data, this 3 times. Then we give the tape back to the vendor.
On about 1200 tapes, we have 1-2 write errors every day. Tapes made by
Imation (ex 3com), not so good!
Yours,
René LAMBELET
NESTEC SA
Information Technology
Av. Nestlé 55 CH-1800 Vevey (Switzerland)
tél +41 (0)21 924 35 43 fax +41 (0)21 703 30 17
mailto:rene.lambelet AT nestle DOT com
-----Original Message-----
From: Pole, Stephen [mailto:Stephen.Pole AT HEALTH.WA.GOV DOT AU]
Sent: Tuesday,29. April 2003 11:08
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: When to declare a tape dead
Hmm depending on your budget and how valuable the clients data is I guess.
And how much "nerve" you have. :-))
I reckon, 1 write error is enough for me to move the data and get that
data
across to a new tape.
Check the contents and verify the tape. Then move a.s.a.p
Cheers
Stephen
-----Original Message-----
From: Adam J. Boyer [mailto:Adam.J.Boyer AT FRB DOT GOV]
Sent: 21 April, 2003 9:57 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: When to declare a tape dead
Greetings all,
We're currently using a 3494 with 3590E drives, and
I have accumulated several tape volumes with write errors = 1, and one with
write errors > 1. Just wondering how I should determine if a tape is
truly dead or "mostly-dead", and when a given error is ignorable or
workable? Is the number of write errors a good indication?
(I've noticed that the errors are distributed across drives, so as not to
point to a single drive. Also, all drives have been recently upgraded to
new microcode.)
Thanks much,
adam
pic14589.jpg
Description: JPEG image
|