ADSM-L

[no subject]

2015-10-04 17:50:03

I *think* that OMI's advantape device driver and IBM's device driver for the Quantum DLT 7000 use different methods for calculating CRC or block checksums. The first two conversion attempts, we managed to get the 9710 and the tape drives defined but an audit volume of an existing tape would give us errors and all files would be defined as unreadable.

I just knocked together a test server yesterday.

It turns out that I need to use the OMI driver, with the associated tape drives (defined as /dev/rmtx) defined as type 'generictape' to read the tapes. I was able to define the same drives, with the ADSM driver (/dev/mtx) as type 'dlt' and move data to them.

It's ugly (I *should* be able to read the tapes with *no* mickey-mouse fix) but at least I can move the incremental data over to the 'new' format and let the archives age off gracefully.

Thanks, Bob.

TomK



    -----Original Message-----
    From:   Bob Miller [SMTP:CBRM AT ARGUSHEALTH DOT COM]
    Sent:   Tuesday, December 08, 1998 6:36 PM
    To:     ADSM-L AT VM.MARIST DOT EDU
    Subject:        Re: Conversion to ADSM version 3

    I am not real familiar with the STK 9710 but could it need a later Microcode version for the hardware?  Our 3490s needed to be upgraded when we went to 3.1.

    >>> "Kauffman, Tom" <KauffmanT AT NIBCO DOT COM> 12/07/98 03:01PM >>>
    I'm having a REAL problem getting to ADSM version 3 -- any suggestions
    greatly appreciated!

    We're currently running ADSM V2.1.5.13 ('enhanced') on AIX 4.2.1 with
    Storage Tek's ACSLS version 5.1 and an STK 9710.

    I have been trying to go to ADSM version 3.1.x *without* the STK
    software. ADSM installs clean, can find the STK library -- but can't
    read the current tapes. The most recent 'fix' was to install the latest
    version of the 'enhanced' ADSM and migrate from there. This doesn't look
    like it will work either.

    I have no problem with a mixed definition for a short time period - the
    vast bulk of my data is in archive format that ages off over either 8 or
    21 days - but I've *got* to be able to read it until it goes away. The
    600 GB of incremental data can be copied as time permits.

    I'm finally in a position to try things out in a test environment, and
    will be re-opening a trouble call with IBM tomorrow - but if there are
    any ideas here, I'd appreciate them.

    Tom Kauffman
    Sr. Technical Analyst
    NIBCO, Inc.

    kauffmant AT nibco DOT com
                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                   !

------ =_NextPart_001_01BE2374.54FE150E-- =======================================================================
<Prev in Thread] Current Thread [Next in Thread>
  • [no subject], Unknown <=