Hello,

did you also notice that the memory problem is not resolved?

I checked the Netware memory and coud find that:

after loading the dsmc schedule, adsm has a size of 250 KB.

After a backup, the size stays at 1.25 MB instead of being released!!

Any comment form IBM?

Regards,

Rene Lambelet
Nestec SA - 55, Av. Nestle - CH-1800 Vevey
Tel: ++41'21'924'35'43 / Fax: ++41'21'924'45'89
E-Mail: rene.lambelet AT nestle DOT com



> -----Original Message-----
> From: Paul Zarnowski [SMTP:vkm AT CORNELLC.CIT.CORNELL DOT EDU]
> Sent: Thursday, June 10, 1999 2:45 AM
> To:   ADSM-L AT VM.MARIST DOT EDU
> Subject:      NetWare performance w 3.1.0.7 client and older server
>
> This is a PTF7 NetWare performance question:
>
> Does anyone know if there are any known performance problems running the
> new 3.1.0.7 Netware client with a server earlier than 3.1.2.30?  The PTF7
> client adds LFN support (yay!), but the README file indicates that the
> server needs to be at 3.1.2.30 to get the LFN support.  Here's why I ask
> about performance..
>
> Our server is still at 3.1.2.13.  One of our Netware users upgraded their
> client to PTF7.  That evening (2 nights ago), and the evening after (last
> night), their backups slowed waay down.  Like from 3-4 hours before
> upgrading to 18+ hours after upgrading.  During the first incremental, all
> the files on only 1 of their volumes were "updated" (according to the
> scheduler log).  Only 1 volume - none of the other volumes - did this.
> This updating took a long time.  Besides that, the processing of files on
> other volumes also slowed way down.  This is for files that did not
> change.
>  You know, there are lines in the scheduler log indicating "nn,nnn files
> processed"?  The timestamps on these were MUCH further apart on the PTF7
> client, than they were before they upgraded, indicating that it was taking
> ADSM much longer just to process files.  I should state that the node was
> upgraded from a version 2 client.
>
> I can schedule an upgrade of our server, but I'd like to know if this is
> likely to solve the performance issue or if the problem lies elsewhere.
> Any thoughts from you experts out there?
>
> Thanks in advance.
> ..Paul