ADSM-L

Re: ADSM bugs and support.

1995-10-23 13:06:11
Subject: Re: ADSM bugs and support.
From: Sheelagh Treweek <sheelagh.treweek AT COMPUTING-SERVICES.OXFORD.AC DOT UK>
Date: Mon, 23 Oct 1995 17:06:11 +0000
If you are happy with ADSM and it meets your needs, please delete now -
apologies for a long post ...

On Mon, 23 Oct 1995, Helmut Richter wrote:

>> I would like to ask ADSM users if many are happy with a product or support.

>This question has got mainly positive answers on this list which squares
>well with the observation that most list participants have a mainframe
>based ADSM installation for backing up real small systems. We, however,
>use ADSM for a couple of larger workstation clusters and are less than
>happy with it:

< ... lots cut ...>

I regret to confirm that here at Oxford University we are extremely
disappointed with the way that ADSM is shaping up (or not, depending on
your viewpoint).

Background:
----------
We bought a twin AIX server solution - to run ADSM on one R24 and to have
We bought a twin AIX server solution - to run ADSM on one R24 and to have
ftp/dfs/nfs with a migrating filesystem on the other R24. Each R24 has 512MB
memory. We have a 3494 with 4 3590 tape decks and 170GB disk storage between
the two systems. We wish to run a site-wide backup/archive scheme for
potentially 10000 users as well as extended filestore facilities.

Additionally, we would very much like to implement a more intelligent
store/search with an integral metadata facility for data such as manuscript
images - potentially terrabytes of data.

All of this we believed  possible (though not all available today).


Problems:
--------
In reality, ADSM struggles to cope with new backups when only 100 nodes are
In reality, ADSM struggles to cope with new backups when only 100 nodes are
registered. Last week, adding just 4 servers (10GB in all) took almost 12
hours with an aggregate 0.25MB/sec for migration between disk and tape. It
would seem that simultaneous client backup and migration from disk pool to
tape pool is not feasible. We already have 8Gb for the main primary
backuppool; we have other storage hierarchies as well - making use of multiple
domains etc. Nice features which we believed were necessary to implement a
complex storage management solution.

I have reported quite a few problems with ADSM 2.1 during the last month; I
won't bore you with all of them but will highlight just two:

(1) When executing about 30 "delete volumes" to facilitate getting rid
of the 3490 file references in ADSM. ADSM accepted the processes; it couldn't
cope and was gridlocked. Now I have been told not to do this as it puts a
high load on ADSM, much better to submit one, watch the process/activity log
for completion and then put in the next one. The problem is closed. If I wish
I can submit a design request change via sales. Of course, I now have a
procedure which accomplishes this task ... but what if 30 different
administrators had individually put in such requests? Do they have to
telephone each other first and come to an arrangement?

(2) Another example of one I am getting nowhere fast with is the Solaris
client failing to perform "prompted" scheduled sessions. This worked at ADSM
1.2; two clients that we updated to ADSM 2.1 immediately failed with messages
like:

19-10-1995 03:00:01      ANR8212W Unable to resolve address for <junk chars>.
19-10-1995 03:00:01      ANR2716E Schedule prompter was not able to contact
                          client
                          OXMAIL3 using type 1 (<junk chars> 53794).
19-10-1995 03:01:08      ANR8212W Unable to resolve address for <junk chars>.


I have been informed that this cannot be reproduced by IBM, with the
implication that this is my fault. I have now reproduced this problem by
bringing two more solaris clients into ADSM - evidence sent to CALL-AIX
today.

Conclusions:
-----------
I believe that the philosophy behind ADSM is excellent and it is now being
I believe that the philosophy behind ADSM is excellent and it is now being
sold to accomplish some quite ambitious projects (like ours).

I am looking for some reassurances by IBM that we have been sold an
appropriate product for our requirements.

When problems are reported, with evidence, which takes a considerable amount
of time to collect/send/chase up, please appreciate that these are real
problems to us and that if the PMR is closed without a resolution or even a
hope for the future, just what are we supposed to do?

When can we look forward to a version of ADSM that will meet the sort of needs
that I have described?

Thanks for your time, Regards,

--
Sheelagh Treweek                         Email: sheelagh.treweek AT oucs.ox.ac 
DOT uk
Sheelagh Treweek                         Email: sheelagh.treweek AT oucs.ox.ac 
DOT uk
Oxford University Computing Service      Tel:   +44 (0)1865 273205
13 Banbury Road, Oxford, OX2 6NN, Uk     Fax:   +44 (0)1865 273275
------------------------------------------------------------------------------
=========================================================================
<Prev in Thread] Current Thread [Next in Thread>