ADSM-L

Re: NT server restore delay

1999-12-21 11:28:06
Subject: Re: NT server restore delay
From: Nathan King <nathan.king AT USAA DOT COM>
Date: Tue, 21 Dec 1999 10:28:06 -0600
David,

I see this behaviour also and have attributed it to the same cause.

Data might not necessarily be contigous on the tape. That's why reclamation
exists.

The data was written to the tape contigously but as files are updated i.e.
actives becoming inactives, I would expect that data would cease to be
contigous.

As a test you could perform a selective backup and then do a restore of all
active files (since the selective should cause all active data to be written
contigously to tape). If you're still getting a lot of these pauses then
we'll have to come up with some other theory.

Nathan

        -----Original Message-----
        From:   Irvine, David (ANFIS) [SMTP:dirvi1 AT ANFIS.CO DOT UK]
        Sent:   Tuesday, December 21, 1999 9:46 AM
        To:     ADSM-L AT VM.MARIST DOT EDU
        Subject:        NT server restore delay

        This may be a totally silly question but . . .

        During a restore of an NT server. I noticed that the restore would
"pause"
        for 2 or 3 minutes at a time. Carrying out a "Query Session" command
shows
        that the same tape is still mounted and the Sess State is "Run"
whilst the
        Wait Time is "0s".

        The delay is therefore not due to tape mounts nor Send Waits due to
a slow
        client. The Tapes in question are 9840 and can contain upwards of
50Gb. Is
        this delay due to the tape being rewound or forwarded as ADSM looks
for the
        client data ?

        I always thought that client data was contigous on a given tape. . .

        David Irvine
        Technical Specialist, IT Services
        Abbey National Financial and Investment Services plc
        Abbey National House 287 St Vincent Street Glasgow G2 5NB
        Registered in Scotland No 159852

        EMail: dirvi1 AT anfis.co DOT uk
        Tel : +44 (0)141-275-7731
        Fax : +44 (0)141-275-9199
<Prev in Thread] Current Thread [Next in Thread>