cancel
Showing results for 
Search instead for 
Did you mean: 

Warning - drive index 0 usage has exceeded the Cleaning Frequency

ipmanyak
Level 5

In windows system log  every 5 min I see this message.

I have two tape drive. Cleaning Frequency  is set 120 and 140 hours

Screenshots  are in the  attachments.

And how to know what drive have index 0 ?  One tape drive  is tape library  (robotics library)  for 8 tapes,  but  cleaning tape in it  is out.

Any ideas ?

2 ACCEPTED SOLUTIONS

Accepted Solutions

sdo
Moderator
Moderator
Partner    VIP    Certified

Some cleaning tapes are good for 20 cleans, and some are good for 50 cleans, and varying amounts in between.

Is a cleaning tape present in the library?

Cleaning tapes do not last forever.  All cleaning media only ever a specific cleaning count.  And, obviously, one mount = one clean.

So, has the cleaning tape expired it's usage count?

.

FYI - all LTO media (all types, all variants, all capacities, all families, all vendors) all have a tiny 'near field' (RFID) chip in them that the LTO tape drives query and update (N.B: this has nothing to do whatsoever with NetBackup) - and so tape drives can detect how many mounts any tape, including cleaning tapes, have ever had, and so tape drives know when a cleaning tape has expired, i.e. tape drives when a cleaning media has hit its maximum clean usage count...

...and so, when a tape drive requests a clean, and a cleaning tape gets mounted, then the tape drive checks the tape, and if it has had too many mounts/uses/cleans, then it simply dismounts the tape without loading it, and so the tape drive does not get cleaned.

.

Anyway - your problem may be down to a mixed mis-configuration of cleaning regime.  If your library is taking care of cleaning then you do not need to enter a cleaning interval of 120 or 140 hours.  You only enter a cleaning interval into NetBackup if all of these following points are true:

1) The library cannot, or does not, or you do not want the library to, manage cleaning.

2) NetBackup cannot detect SCSI tape alerts for when tape drives request a clean.

3) You want to manually override cleaning alerts and force cleaning no matter whether the tape drive has requested a clean or not.

.

Interval based cleaning is usually not a good idea, nor (in most cases) recommended.  Most tape drive vendors do not recommend interval based cleaning - and instead recommend alert based cleaning - because the cleaning tape is a form of mild abbrassive - and each act/event of cleaning has to attempt to scratch dirt (oxide accumulation) off of the tape drive head, but not so roughly that it scratches the tape head itself.  In short, too many, too regular, cleans... shortens the life span of tape drive read/write heads - and thus slowly increases the chances of read and write failures.

.

If you don't know the answers, or don't understand any of the above, or you are unable to describe to yourself what your cleaning regime is, and more importantly if you do not understand why your cleaning regime is configured the way it is - then you need to start from the beginning, and read the manuals and release notes for the tape library and for the drives and read the relevant sections of the NetBAckup admin guides - and have a plan - before making any other changes.

View solution in original post

Marianne
Level 6
Partner    VIP    Accredited Certified

Show us output of 'tpconfig -l'  and 'vmoprcmd -d'.
(Commands are in ...\veritas\volmgr\bin)

Best NOT to configure periodic cleaning and rather rely on TapeAlert for cleaning.

To let NBU know that you have manually cleaned the drive, issue this command (from same dir as above):

tpclean -M <drive_name>

Some reading matter: 

How NetBackup TapeAlert works
http://www.symantec.com/docs/HOWTO99812 

https://www-secure.symantec.com/connect/forums/tape-drive-cleaning-0

 

View solution in original post

12 REPLIES 12

sdo
Moderator
Moderator
Partner    VIP    Certified

Some cleaning tapes are good for 20 cleans, and some are good for 50 cleans, and varying amounts in between.

Is a cleaning tape present in the library?

Cleaning tapes do not last forever.  All cleaning media only ever a specific cleaning count.  And, obviously, one mount = one clean.

So, has the cleaning tape expired it's usage count?

.

FYI - all LTO media (all types, all variants, all capacities, all families, all vendors) all have a tiny 'near field' (RFID) chip in them that the LTO tape drives query and update (N.B: this has nothing to do whatsoever with NetBackup) - and so tape drives can detect how many mounts any tape, including cleaning tapes, have ever had, and so tape drives know when a cleaning tape has expired, i.e. tape drives when a cleaning media has hit its maximum clean usage count...

...and so, when a tape drive requests a clean, and a cleaning tape gets mounted, then the tape drive checks the tape, and if it has had too many mounts/uses/cleans, then it simply dismounts the tape without loading it, and so the tape drive does not get cleaned.

.

Anyway - your problem may be down to a mixed mis-configuration of cleaning regime.  If your library is taking care of cleaning then you do not need to enter a cleaning interval of 120 or 140 hours.  You only enter a cleaning interval into NetBackup if all of these following points are true:

1) The library cannot, or does not, or you do not want the library to, manage cleaning.

2) NetBackup cannot detect SCSI tape alerts for when tape drives request a clean.

3) You want to manually override cleaning alerts and force cleaning no matter whether the tape drive has requested a clean or not.

.

Interval based cleaning is usually not a good idea, nor (in most cases) recommended.  Most tape drive vendors do not recommend interval based cleaning - and instead recommend alert based cleaning - because the cleaning tape is a form of mild abbrassive - and each act/event of cleaning has to attempt to scratch dirt (oxide accumulation) off of the tape drive head, but not so roughly that it scratches the tape head itself.  In short, too many, too regular, cleans... shortens the life span of tape drive read/write heads - and thus slowly increases the chances of read and write failures.

.

If you don't know the answers, or don't understand any of the above, or you are unable to describe to yourself what your cleaning regime is, and more importantly if you do not understand why your cleaning regime is configured the way it is - then you need to start from the beginning, and read the manuals and release notes for the tape library and for the drives and read the relevant sections of the NetBAckup admin guides - and have a plan - before making any other changes.

Marianne
Level 6
Partner    VIP    Accredited Certified

Show us output of 'tpconfig -l'  and 'vmoprcmd -d'.
(Commands are in ...\veritas\volmgr\bin)

Best NOT to configure periodic cleaning and rather rely on TapeAlert for cleaning.

To let NBU know that you have manually cleaned the drive, issue this command (from same dir as above):

tpclean -M <drive_name>

Some reading matter: 

How NetBackup TapeAlert works
http://www.symantec.com/docs/HOWTO99812 

https://www-secure.symantec.com/connect/forums/tape-drive-cleaning-0

 

sdo
Moderator
Moderator
Partner    VIP    Certified

Interval based cleaning can be thought of in at least three ways:

1) A left-over feature/capability of NetBackup for older tape drive technologies such as DDS, DAT, DLT, SDLT, AIT, etc. - which did not implement alert based cleaning, and where the tape drive vendors specifically recommended regular cleaning.

2) A required feature for other (non-LTO) modern tape drive technologies which do not implement alert based cleaning.

3) A required feature for when tape libraries and/or LTO tape drives are unable (for whatever reason) to alert to within the library itself, or even to NetBackup.

.

LTO is typically alert based cleaning, and not interval based cleaning.  But sometimes, our only option is interval based cleaning.  If I have LTO tape drives, I would do my best to avoid interval based cleaning, and do my research and try to implement either library based/managed cleaning (preferably) - or NetBackup based/managed cleaning (second choice) - before looking at the third choice of interval based cleaning.

.

P.S. Take a look at this recent post too:

https://www-secure.symantec.com/connect/forums/cleaning-media-3

sdo
Moderator
Moderator
Partner    VIP    Certified

Usually, the choice of magazine is irrelevant, but then again it might not be - sometimes magazines are restricted to certain sides/spaces slot ranges within a library - and this gets more confusing if tape library partitioning is active.

.

Anyway, was it a brand new sealed (never opened before) cleaning tape?

Or an already opened (unsealed) cleaning tape which may have been used before?  If so, maybe the internal mount count has been exceeded - and maybe the tape drive did not get cleaned, even though the cleaning was mounted (briefly).  Did the cleaning tape stay mounted for a few minutes at least?

Maybe when you initiated a clean of drive 0 - it cleaned the other drive?  Maybe the drive that you think is 'drive 0' is the other drive?

.

I, nor anyone else, can answer what you have to do... it's all down to the combination of tape library features/capabilities, tape drive features/capabilities - and the tape library and/or tape drive 'driver' features/capabilities, and HBA (SCSI, SAS, FC) features/capabilities - and sometimes even down to O/S implementation of two way SCSI traffic - i.e. are tape drives able to send an alert upwards without a SCSI sense/query having been initiated.

And sometimes, the tape library/drive vendor documentation is weak, or non-existent - in which case your only option is to have to try to implement a particular regime in order to discover that it does not work - so that you can try to implement another alternative/different regime.

ipmanyak
Level 5


C:\Program Files\VERITAS\Volmgr\bin>tpconfig -l
Device Robot Drive       Robot                    Drive                 Device
Type     Num Index  Type DrNum Status  Comment    Name                  Path
drive      -    0 hcart2    -      UP  PLEHANOVO  HP.ULTRIUM5-SCSI.000  {5,7,0,0}

C:\Program Files\VERITAS\Volmgr\bin>tpclean -M HP.ULTRIUM5-SCSI.000


C:\Program Files\VERITAS\Volmgr\bin>vmoprcmd -d

                                PENDING REQUESTS

                                     <NONE>

                                  DRIVE STATUS

Drv Type   Control  User      Label  RecMID  ExtMID  Ready   Wr.Enbl.  ReqId
  0 hcart2   AVR               Yes   A00026  A00026   Yes     Yes        0

                             ADDITIONAL DRIVE STATUS

Drv DriveName            Shared    Assigned        Comment
  0 HP.ULTRIUM5-SCSI.000  No       tmn-backup.utai PLEHANOVO

--------------

C:\Program Files\VERITAS\Volmgr\bin>tpclean -M HP.ULTRIUM5-SCSI.000

was run without any messages

---------------

Cleaning tape was run in robotics  library 1 min.

 

 

ipmanyak
Level 5

I found that drive  index 0 - is tape library  (robotics library) and  in Drive detail Information  in Cleaning  Comment  - "Need Cleaning".  I removed  Right Magazine, inserted cleaning tape  and  make cleaning manually from  display panel  successfully. But  then in  Drive detail Information  in Cleaning  Comment I see again  - "Need Cleaning".  Why ?

Do I have to establish cleaning frequency in zero  for my robotics library?

ipmanyak
Level 5

sdo, my cleaning tape is  a new, count not exceeded 10 mount. Earlier  when count of cleaning tape   was exceeded  my tape drive  gave out failure and cleaning not working.

Now I see drive information   of both tape drive,   there both have an index 0 

 

C:\Program Files\VERITAS\Volmgr\bin>tpconfig -l
Device Robot Drive       Robot                    Drive                 Device
Type     Num Index  Type DrNum Status  Comment    Name                  Path
drive      -    0 hcart2    -      UP  PLEHANOVO  HP.ULTRIUM5-SCSI.000  {5,7,0,0}

C:\Program Files\VERITAS\Volmgr\bin>vmoprcmd -d

                                PENDING REQUESTS

                                     <NONE>

                                  DRIVE STATUS

Drv Type   Control  User      Label  RecMID  ExtMID  Ready   Wr.Enbl.  ReqId
  0 hcart2   AVR               Yes   A00026           Yes     Yes        0

                             ADDITIONAL DRIVE STATUS

Drv DriveName            Shared    Assigned        Comment
  0 HP.ULTRIUM5-SCSI.000  No       -               PLEHANOVO

C:\Program Files\VERITAS\Volmgr\bin>tpclean -M HP.ULTRIUM5-SCSI.000 - was run without any messages

 

ipmanyak
Level 5

I run command  for my robotics library

tpclean -M HP.ULTRIUM3-SCSI.000

and then the message in Cleaning Comment - "Need Cleaning"    disappeared.  :)

So you suggest to remove the  Сleaning frequency parameter on all tape devices?  And for this purpose to establish this parameter in zero? 

 

sdo
Moderator
Moderator
Partner    VIP    Certified

Apologies, I'm not suggesting to reduce the frequency to zero - I was only trying to outline why sometimes it may not be appropriate - I can't tell you what is, or what is not, approrpiate for your environment - because so many factors come in to play - it's just not possible for me to comment/suggest/recommend - other than to do some research, and planning, and possibly some testing.  You could well leave the environment as it is, and never have a problem again.  Who knows?  All I wanted to do was to steer your thought process somewhat, to help you consider the possibilities, capabilities, limitations.  All of which I have no visibility of.  Only you can make the required decision based upon an informed and factual position.

Marianne
Level 6
Partner    VIP    Accredited Certified

Have you gone through the TN and forum discussion that I have posted above?

Best to disable Frequency cleaning and allow the tape drive to request cleaning via TapeAlert.

ipmanyak
Level 5

Yes Marianne. I read it.

Nicolai
Moderator
Moderator
Partner    VIP   

Do you have SCSI sence key codes in event viewer. I suspect you may have a bad tape or drive that trigger the clean flag.