ADSM-L

Re: TSM and device block sizes

2000-05-29 08:55:09
Subject: Re: TSM and device block sizes
From: Richard Sims <rbs AT BU DOT EDU>
Date: Mon, 29 May 2000 08:55:09 -0400
>Richard-well actually I wasn't being entirely honest about the
>documentation!  I read about the fixed blocksize in a PMR on RETAIN.
>lsattr shows me block size of 0.  I can, of course change it, but what I
>read indicates it gets ignored.

Phil - I'm glad to see that it's 0, which is what one should expect.
       That allows *SM to optimally use the best blksize for the drive.
*SM is setting options itself, for best use of the hardware, and we should
not try to interfere with that.  If you were really curious about the
block sizes that ended up on the tape you could use an independent tape
utility to scan the tape and glean that info.
    In general, though, building a paradigm on top of hardware so as to
achieve some kind of logical advantage results in a performance compromise,
as in this kind of device emulation.  Using a device natively always
yields the best performance; but you don't have that choice within the
given implementation.
      Richard Sims, BU
<Prev in Thread] Current Thread [Next in Thread>