Hello,
If you read in the 3494 Operator's Guide, it mentions that the "performance
of the drive" is monitored to determine when to clean it, which I believe is
done by measuring voltages across the head. For most TSM servers, this
monitored cleaning will occur much sooner and more often than the scheduled
cleaning defaults.
So, Zoltan, you should not have to adjust the scheduled cleaning option. If
the current setting is to clean the drives every 65 mounts, you may be
seeing those cleanings in addition to the "as needed" cleanings initiated by
the LM. I would be tempted to set it back up to 999, and see if the
cleanings are reduced without causing any TSM I/O errors. Check with your
CE first though - maybe he/she knows the history of why it was reduced in
the first place. Perhaps it was to address a problem with older technology
drives (i.e, 3490).
Milton, that's a clever idea to keep an eye on the cleaning history - I've
back-tracked that information often when trouble-shooting drive problems.
Knowing it beforehand would be an excellent way to predict failures.
- Paul
-----Original Message-----
From: Johnson, Milton [mailto:milton.johnson AT CITIGROUP DOT COM]
Sent: Friday, June 04, 2004 2:10 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: 3590 cleaning scheduling
"Clean Me" requests are also generated when the drive encounters a
"problem". A faulty/failing drive can cause an abnormal spike in
cleaning requests. I monitor these requests daily as an indicator of
drive health. I have found that an increase in "Clean Me" messages
correlates with an "Attn Drive" message on the LCD of the 3590E1A.
-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf Of
Zoltan Forray/AC/VCU
Sent: Friday, June 04, 2004 1:10 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: OT?: 3590 cleaning scheduling
Lately, I have been noticing lots of "Clean Me" messages in ERRPT of the
TSM AIX system, for the 3590E1A drives it uses.
So, I started looking into the setting on the 3494, which controls
cleaning. Currently, it is set to clean by usage (65).
Upon digging through the "IBM TotalStorage Enterprise Tape - A Practical
Guide" Redbook, I found the paragraph:
We recommend that drive cleaning be based on a drive request. Use
time-based cleaning only if drive usage is very low. For 3590 tape
drives, use a value of 999 mounts to perform cleaning based on a drive
request rather than an initiated library.
So, what do you use to control the cleaning cycle for your 3494/3590
drives ? Do you follow the above recommendation ? Is this a global
recommendation that can be applied to *ALL* 359x drives ?
Inquiring minds want to know ?
|
