Conference ssag::ask_ssag

    The switches available on the SBB and the switches available on the
    TLZ07 or TLZ7L are completely different switches, and have -nothing- to
    do with each other. Nothing.
    
    The docs for the -VA are accurate. The docs for the drive are also
    accurate, except for the minor detail that -none- of the drive switches
    are accessible from the outside of the SBB. They don't need to be.
    
    The drive switches set drive defaults. O'VMS overrides those defaults.
    It is highly unlikely that internal drive switch settings have anything
    to do with your problem. 
    
    We've been shipping StorageWorks for years now - the differentiation
    between SBB switches and drive switches has been beaten to death in the
    docs, training, and notes for all that time.... 
    
    Chris
    

    Thank you for your prompt and authoritative reply, and apologies
    for my ignorance. We in the field tend to know and learn what we
    have to, and until quite recently I have needed to be an RA82/
    TA78 guru; now I have TWO TLZ7L's in my account base. Times are
    changing! (Now if I could get rid of that DECService LP14...)
    
    CR

    Well, I just returned from the customer site and I'm afraid I must
    disagree with some of Mr. Williams' reply (.1). My new replacement 
    behaved the same way the original replacement did, so we were forced
    to experiment (as we often are in the field.) The drive works groovy
    now. re: 6527.1:
    
    "The docs for the -VA are accurate..."
    	(yeah, as far as they go.)
    "The docs for the drive are also accurate, except for the minor detail
    that -none- of the drive switches are accessible from the outside of
    the SBB..."
    	(Look again. An 8-switch DIP pack on the drive itself, not as sub-
       	miniature as the SBB switches, plainly visible and configurable.)
    "They don't need to be."
    	(..unless you want data compression. Read on.)
    "O'VMS overrides those defaults."
    	(Not by any means or agency the customer and I were able to find.
   	 The old drive compressed, my spare didn't. No S/W changes.)
    "It is highly unlikely that internal drive switch settings have
    anything to do with your problem."
    	(SET *DRIVE* SWITCH 6 OFF and enable compression...or disable lack of
   	 compression if you prefer.) {See .0 on ergonomics.}
    	(It may be "highly unlikely" but it fixed MY problem.)
    "the differentiation between SBB switches and drive switches has been
    beaten to death in the docs, training, and notes..."
    	(Maybe so, but nowhere in the docs does it say that if you buy this
    data-compression-capable option, the default setting is compression
    DISABLED, and to get what you paid for you must set a switch which you
    are not supposed to touch. Or even look for.)
    
    It appears to be the basic "unless you mean" situation: never touch the
    drive switches on an SBB, unless you mean the compression disable
    switch, which must be set OFF if you want data compression. I would
    address these questions offline except that Mr. Williams' counsel was
    made available to all, and it appears to me that some of it may be
    incorrect. I hope that by staying with the ASK_SSAG forum we can avoid
    some problems in the field.
    
    Chris Reed MCS East Providence, RI

    Since the first time we put Compression into the DAT drives, OVMS has
    overridden the default setting on directly connected busses.
    The only exception is when connected to HSx controllers (which, by the
    way, is an unsupported configuration, per the HSx folks...), where
    compression has to be set from the HSx console.
    
    If it is not doing that, OVMS has a bug. Check for patches - OVMS has
    had a lot of trouble getting the compression control code implemented.
    If no patches are available for your version, IPMT it.
    
    I stand by all the comments in my previous note. They are accurate,
    when OVMS is working the way that it is supposed to. If the switches on
    the TLZ7L drive happen to be accessible from the outside, it's not by
    design. You are not supposed to ever have to get to them. 
    
    You -are- using the /Media_format=compaction switch, no? On Init,
    Mount, Backup, etc?
    
    Chris Williams
    Tape and Optical Engineering 
    ex- TLZ04, TLZ06/6L engineer
    

    I'm confused to where the compression switch is in this discussion but
    it no /MEDIA is specified OleVMS uses CURRENT compression. To really be
    sure about compression the BACKUP command should include /MEDIA=COMPACT 
    AND /REWIND. 
    
    Guenther

    .4 has the key here. The TLZ7L in question is NOT on a directly-
    connected bus. It is on an HSD10. As it says in .2, "Read ALL of them."
    I didn't have the imagination or experience to know that compression
    was affected when the STGWKS device is not on a native bus.
    
    Fortunately, setting Switch 6 OFF seems not to have broken anything. I
    will talk with the customer and schedule one more session to RE-set the
    switch and make changes to the HSD setup parameters; I think he will
    accede to this in the interest of uniformity, especially since he can
    use the current setup as a fallback in case there are even more factors
    I've overlooked. Since the H/W config is unsupported (sold, but
    unsupported) I will be alert for funny behavior.
    
    For the record, I did search this notesfile for the TLZ7L-VA and
    COMPRESSION strings, but I only went back to Note 4xxx. (I've only known
    of ASK_SSAG for a few months, and with 6500-plus notes I have a lot of
    catching up to do. Sorry to have shed heat vs light on the situation.)
    
    CR

    re: -.1:
    
    where I said ".2" I meant ".1". Also, yes, we used the proper commands
    and qualifiers to invoke Compaction. They work as intended now with
    hardware switch 6 on the back of the drive OFF and the switch pack on
    the back of the SBB set up per the Storageworks leaflet shipped with
    the drive (SCSI target 5 regardless of Storageworks slot, in our case.)
    
    As I mentioned above I will lobby the customer to try to do this with
    the standard H/W config in order to reduce the chance of a similar
    adventure on an afterhours call with a different F.E.

    One other caveat on all this: While with many tapes OVMS will use the
    current compression state of the tape to do appends, this is -not- the
    case with any of the DAT tapes, at least on direct connects. The OVMS
    drivers don't seem to ask the drive about the compression state. If
    compression is specifed, it compresses. If not, it does not. Omitting
    the switch defaults the drive back to uncompressed.
    
    This can lead to some rather interesting mixed
    compression/non-compression tapes. The drive could care less, however
    (by design!) - it will just decompress anything that needs to be on
    reads. Best to physically mark the tapes, if you want to avoid this...
    
    Chris