ADSM-L

Re: [ADSM-L] 5.4 performance issues

2008-03-17 16:29:55
Subject: Re: [ADSM-L] 5.4 performance issues
From: Michael Prix <michael.prix AT RS6000.DARKTECH DOT ORG>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Mon, 17 Mar 2008 21:27:53 +0100
Dave,

it's not only the 5.4 client code! We have to use 5.5-client code and having
the same performance issues.

Client: 5.5.0
Server: 5.4.2
OS: Linux 64-bit

--
Michael Prix


>          IBM has had several customers reporting this performance issue
> after upgrading to level 5.4. APAR IC53531 was originally opened for this
> for the NetWare platform only. We are now also seeing the same issue for
> this APAR for customers on UNIX platforms as well. (However, this APAR does
> not apply to Windows platforms.) For Netware, the interim fix is 5.4.1.4. I
> do not have the date for the UNIX platforms yet.
>
> At 05:02 PM 3/14/2008 -0400, you wrote:
> >Hi All,
> >
> >      We recently upgraded the TSM clients on our cryus imap servers from
> >5.3.0.12 to 5.4.1.2 (and then 5.4.1.5), and since that upgrade, we have
> >seen the backups of about 4.5million files in each of our 12 spools go
> >from about 2.5 hours to approx. 4 days (had we let it run to completion).
> >
> >Anyone else seeing issues with 5.4.x on Solaris clients with millions of
> >files?
> >
> >Other vitals:
> >  Server: 5.4.1.2 on AIX
> >  Network: GB
> >
> >  Client OS Solaris 2.9
> >   Filesystems are VxFS on an Hitachi 9960.
> >
> >We have opened a PMR with IBM, and they want to run some traces, but we
> >have since reverted back to 5.3.0.12, and our 2.5hr backups.
> >
> >We need to run 5.4 to stay supported, and a minimum of 5.4.1.2 for the
> >fix for the security issues in dsmcad, which we use for email restores on
> >these systems.
> >
> >Thanks in advance for any insights anyone can provide.
> >
> >
> >Steve Roder
> >University at Buffalo
> >(spr AT buffalo DOT edu | (716)645-3564)
>
> Dave Canan
> TSM Performance
> IBM Advanced Technical Support
> ddcanan AT us.ibm DOT com

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