Re: How many "tape media errors" is cause for concern?

[Jim Oberholtzer <midrangel@xxxxxxxxxx>]

Changes in commands won't cause the kind of issues your having.

I know you use Fibre,   Are you sure the fibre cables in good shape?  
Did someone put a Size 12 crimp in them? (or what ever size shoe you 
care about)   Recheck the fibre cables.  Did someone mess with one of 
the Fibre switches, of course nothing was done that would have affected 
the IBM i partition would it?

Barring all of that, I'll bet the interface card has a new part number, 
and new microcode.

Jim Oberholtzer
Chief Technical Architect
Agile Technology Architects


On 1/26/2012 7:35 AM, rob@xxxxxxxxx wrote:
> Argh!  Yet another media error last night - different lpar.  Refreshing
> data gathering to open ticket.
>
>
> Rob Berendt
> -- Group Dekko Dept 1600 Mail to: 2505 Dekko Drive Garrett, IN 46738 
> Ship to: Dock 108 6928N 400E Kendallville, IN 46755 
> http://www.dekko.com From: rob@xxxxxxxxx To: Midrange Systems 
> Technical Discussion <midrange-l@xxxxxxxxxxxx>, Date: 01/25/2012 10:54 
> AM Subject: Re: How many "tape media errors" is cause for concern? 
> Sent by: midrange-l-bounces@xxxxxxxxxxxx Maybe I'll copy a bulk of 
> this email into a new pmr ticket and see what flies. The operator is 
> concerned that another change that was done (among many) when this 
> started occurring was that we changed from native save commands to 
> BRMS. I don't want to start that finger pointing. We're fiber, not 
> pins. I hear you though. After a pi... contest with IBM over bent pins 
> I no longer change cables in the other city with those kind of cables. 
> "Told" the boss to get enough physical drives so that we didn't have 
> to swap cables during downtime to physically share drives. The thought 
> of only being allowed to physically back up once a quarter and have 
> the freaking tape drive go bad because of a bent pin during that time 
> didn't set well with me. As far as I/O cards go, I'll document their 
> part numbers and what not for the pmr and leave that up to IBM. Rob 
> Berendt
> -- Group Dekko Dept 1600 Mail to: 2505 Dekko Drive Garrett, IN 46738 
> Ship to: Dock 108 6928N 400E Kendallville, IN 46755 
> http://www.dekko.com From: Jim Oberholtzer <midrangel@xxxxxxxxxx> To: 
> Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxx>, Date: 
> 01/25/2012 10:41 AM Subject: Re: How many "tape media errors" is cause 
> for concern? Sent by: midrange-l-bounces@xxxxxxxxxxxx Rob, I start to 
> get concerned right away. If the tape device and library firmware are 
> up to current, consider the interface card. In the last month I have 
> had two customers where the tape libraries were just behaving poorly. 
> Sometimes they work, other times, not so much. After finally getting 
> in touch with the correct resource in Rochester, I managed to get them 
> to consider the interface card. It was operational, so no problem 
> right? BUT, the part number was very different from the current one 
> (That's part number not feature code). Every time IBM puts new 
> firmware on a feature code card, they do change the part number to 
> signify a change. The CEs have a cross reference. I convinced them to 
> swap the interface card with a newer part number and shazam, no 
> problems anymore. It might be an incompatibility between the interface 
> card and the tape device. BTW for grins I had them change the cable 
> too. It just might have been part of the problem with moving around in 
> the cabinet. Those ends to tend to bend just a bit, causing unknown 
> problems. Jim Oberholtzer Chief Technical Architect Agile Technology 
> Architects On 1/25/2012 8:56 AM, rob@xxxxxxxxx wrote:
> > >  If your backup aborts with
> > >  CPF5349 - Media error on volume BR0204 on device TAPKVL01. 17:53
> > >  when do you get concerned?  For example, if I had one in 5 years I
> > >  probably wouldn't care.  But at what point do you contact IBM and ask if
> > >  we need to look and see if perhaps something needs to be looked at,
> like,
> > >  perhaps, new drive firmware?
> > >
> > >  3576 (aka TS3310)
> > >  3576-MTL Firmware Version       587G.GS003
> > >  BSP Level       6.86
> > >  All 6 LTO-3 Fibre drives are at 93GE.
> > >
> > >  Tape volume id's differ.  When the tape is next put back into
> circulation
> > >  it is initialized with a CLEAR(*YES) option and always passes.  Multiple
> > >  lpars attach to this via SAN.  Due to zoning and whatnot I am not sure
> if
> > >  it's a particular drive in the library.  Four lpars of i use 3 of the 6
> > >  drives.  One lpar of Linux uses the other 3 drives.  (TSM on Linux uses
> > >  multiple tape drives for deduplication, etc.)
> > >
> > >  Entry     Major                    Major Minor
> > >  ID        Description              Code  Code  Date      Time
> > >  01013B07  Tolerance                0400  0000  01/24/12  17:45:35
> > >  01013B08  Tolerance                0400  0000  01/24/12  17:45:35
> > >  01013B09  Tape support             2E00  0304  01/24/12  17:53:21
> > >  01013B0A  IOP support              3400  0015  01/24/12  17:53:21
> > >  01013B0B  Source/Sink              0700  0120  01/24/12  17:53:39
> > >  01013B0C  Source/Sink              0700  0195  01/24/12  17:53:39
> > >
> > >
> > >  MAILTWO lpar
> > >  Entry     Major                    Major Minor
> > >  ID        Description              Code  Code  Date
> > >  01002C92  Tape support             2E00  0100  11/16/07
> > >  01002C93  Tape support             2E00  0100  11/16/07
> > >  01002C94  Tape support             2E00  0591  11/16/07
> > >  0100407F  Tape support             2E00  0302  03/20/08
> > >  01005C8E  Tape support             2E00  0302  08/19/08
> > >  01010041  Tape support             2E00  0E00  03/31/11
> > >  0101199F  Tape support             2E00  0E00  08/11/11
> > >  01011C2F  Tape support             2E00  0304  08/26/11
> > >  010123D1  Tape support             2E00  0304  10/13/11
> > >  010129EE  Tape support             2E00  0304  11/09/11
> > >  0101337F  Tape support             2E00  0304  12/12/11
> > >  01013B09  Tape support             2E00  0304  01/24/12
> > >  See the change over the years?  MAILTWO uses 2-3 tapes in it's nightly
> > >  backup.  The other 3 lpars use 1 each.
> > >
> > >  GDISYS2 lpar hasn't had an error since '08.
> > >
> > >  GDIHQ2 lpar:
> > >  Entry     Major                    Major Minor
> > >  ID        Description              Code  Code  Date
> > >  01007996  Tape support             2E00  0302  04/30/07
> > >  01007997  Tape support             2E00  0302  04/30/07
> > >  01007998  Tape support             2E00  0302  04/30/07
> > >  01007999  Tape support             2E00  0803  04/30/07
> > >  0100799B  Tape support             2E00  0804  04/30/07
> > >  0100799C  Tape support             2E00  0804  04/30/07
> > >  0100799D  Tape support             2E00  0302  04/30/07
> > >  0100799E  Tape support             2E00  0803  04/30/07
> > >  0100E38B  Tape support             2E00  0302  03/20/08
> > >  01018381  Tape support             2E00  0402  10/01/10
> > >  0101B9D1  Tape support             2E00  0302  06/07/11
> > >  0101D887  Tape support             2E00  0304  09/19/11
> > >  0200298E  Tape support             2E00  8101  12/13/11
> > >  0200298F  Tape support             2E00  0303  12/13/11
> > >
> > >  GDWEB2 contains no errors.
> > >
> > >  VERY SIGNIFICANT SITUATIONAL FACT:  Prior to about the last year we only
> > >  backed these lpars every other month.  As they were HA machines.  Now we
> > >  back them up 5 nights a week and only backup the primary machines once a
> > >  quarter.
> > >
> > >  Rob Berendt
> > >  -- Group Dekko Dept 1600 Mail to: 2505 Dekko Drive Garrett, IN 46738
> > >  Ship to: Dock 108 6928N 400E Kendallville, IN 46755http://www.dekko.com
> > >  -- 
> --