ADSM-L

Re: 40 NT servers suddenly failing

2001-10-25 16:43:56
Subject: Re: 40 NT servers suddenly failing
From: Sung Y Lee <sunglee AT US.IBM DOT COM>
Date: Thu, 25 Oct 2001 15:40:37 -0500
Sounds like network issue.
Depending on how it's configured, but usually in the schedule backup the
TSM server uses  port 1501 to contact client boxes to start the backup.
Sounds like somehow the port 1501 was disabled or turned off.  You can also
test it by telnetting to "ip client 1501" (assuming telnet is not
disabled). If you can't telnet that's your problem.  This is one of many
tests, but seems to work pretty well.

Based on my experience, once the backups miss  the client schedulers need
to be recycled or else they will miss the next day.  Sometimes clients will
work without having to be recycled.   I think it's always good idea to
recycle the schedule after any miss backups.


Sung Y. Lee



                    "Adam J.
                    Boyer"               To:     ADSM-L AT VM.MARIST DOT EDU
                    <Adam.J.Boyer@       cc:
                    FRB.GOV>             Subject:     40 NT servers suddenly 
failing
                    Sent by:
                    "ADSM: Dist
                    Stor Manager"
                    <ADSM-L AT VM DOT MAR
                    IST.EDU>


                    10/25/2001
                    02:45 PM
                    Please respond
                    to "ADSM: Dist
                    Stor Manager"





A few days ago, about 40 NT/2000 clients failed to backup.  The error for
all of them was:

ANR2716E Schedule prompter was not able to contact client
                          CLIENT001 using type 1 (1.2.3.4 1501).

The clients are 4.2.0 and the server is 4.2.1.  Not only did they fail, but
the client scheduler also became useless and had to be stopped and
re-started on all of them.  This had happened in the past to one or two
servers occasionally, but this instance scared me.  I also should mention
that not all of our NT/2000 servers failed, but a lot of them did.

I haven't talked to the network admins yet, but my feeling is that if there
was a network problem, fine, the backups should fail.  But the client
schedulers shouldn't have needed to be stop-started.

Any help would be very much appreciated.


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