ADSM-L

Re: bad performance restoring very small nt-directories

2000-03-30 07:22:38
Subject: Re: bad performance restoring very small nt-directories
From: Richard Sims <rbs AT BU DOT EDU>
Date: Thu, 30 Mar 2000 07:22:38 -0500
>Thank you Richard,
>
>TESTFLAG DISABLENQR resolves my problems.
>
>In which situation should I NOT USE   TESTFLAG DISABLENQR ?
>In which situation should I     USE   TESTFLAG DISABLENQR ?

Andreas - That's great!

I wish I had a good answer for when such an override should be used.
No Query Restore was one of the highlighted new features in the roll-out of
ADSM version 3, one that we all had great hopes would enhance restoral speed
in the way it was promoted.  Unfortunately, it did not work out that way,
instead having the effect of actually worsening the performance of many
ordinary restoral conditions, as you know from your experience.  The body of
APARs addressing problems resulting from NQR and the addition of a developer's
test flag override is further evidence.  I think by this point that NQR is a
sore subject with Tivoli in taking ADSM over from Adstar.

We've all been subject to the NQR problem for some time now; but despite the
substantial adverse effect it can have on restorals, there has been no
statement from Tivoli that I know of as to what's being done to resolve the
problem, or just what customers are supposed to do in the mean time.

Client developers monitor this mailing list, but have thus far been silent as
customers have written in, struggling with the effects of NQR.  We'd really
appreciate someone from Tivoli contributing some information on this.  We have
a real problem here, fellas, and could use some help.

Beyond that, I could only suggest opening a problem report with Tivoli to try
to at least get some information (and please post to the List), or perhaps
ask about it in the next Tivoli teleconference event.  SHARE is also coming up
in a few months; but it would be nice to get info before that.

   Richard Sims, BU
<Prev in Thread] Current Thread [Next in Thread>