ADSM-L

Re: server restore behavior

2006-08-19 14:08:31
Subject: Re: server restore behavior
From: Robin Sharpe <Robin_Sharpe AT BERLEX DOT COM>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Fri, 18 Aug 2006 16:12:36 -0400
AFAIK, it's first come first serve.

- The different restore sessions will not talk to each other, and the TSM
server will not coordinate their resource usage, so it's not A.
- Since each session will compete for drives and get the next one available
when they get up to bat, it's not B.
- And since they all have resourceutilization 4, it's not C.

In fact, I'd think it's quite unpredictable, because each client may have
differing hardware capabilities, other network traffic will influence it,
and what each client is restoring will affect how quickly it gets to the
tape mount request(s).

It would be nice, though, if the TSM server did coordinate the active
sessions.  Even nicer would be a facility to define a restore plan,
assigning priorities and weights.  I suppose you could hack something
together with some fancy scripting and/or using an external scheduler like
Control-M... but seems like it would be a lot of work.

Robin Sharpe
Berlex Labs



             Troy Frank
             <Troy.Frank@UWMF.
             WISC.EDU>                                                  To
             Sent by: "ADSM:           ADSM-L AT VM.MARIST DOT EDU
             Dist Stor                                                  cc
             Manager"
             <[email protected]                                     Subject
             .EDU>                     server restore behavior


             08/18/2006 12:22
             PM


             Please respond to
             "ADSM: Dist Stor
                 Manager"
             <[email protected]
                   .EDU>






This is more of a curiosity question than a problem.  In a
multiple-client restore scenario where you start up say 5 restores at
once from different nodes (with 4 tape drives, and resourceutilization
set to 4 on nodes), how does the server process the request?
Technically, data from all 5 nodes are probably on a lot of the same
tapes.  Does it....

A) mount each tape exactly once, getting all data for all running
restores off that tape before unmounting.

B)  Process the restores relatively serially for each node, giving each
all 4 drives until completed.  Unmounting/remounting the same tapes
multiple times.

C)  Only give each node 1 tape drive to work with, which will
effectively ellicit behavior very similiar to option B.

Or does it do something different than any of these?


Confidentiality Notice follows:

The information in this message (and the documents attached to it, if any)
is confidential and may be legally privileged. It is intended solely for
the addressee. Access to this message by anyone else is unauthorized. If
you are not the intended recipient, any disclosure, copying, distribution
or any action taken, or omitted to be taken in reliance on it is
prohibited and may be unlawful. If you have received this message in
error, please delete all electronic copies of this message (and the
documents attached to it, if any), destroy any hard copies you may have
created and notify me immediately by replying to this email. Thank you.

<Prev in Thread] Current Thread [Next in Thread>