Veritas-bu

[Veritas-bu] Problems when robot was powercycled

2002-10-22 08:49:03
Subject: [Veritas-bu] Problems when robot was powercycled
From: Bruno.Bossier AT comparex DOT be (Bruno.Bossier AT comparex DOT be)
Date: Tue, 22 Oct 2002 13:49:03 +0100
Hello,

We have the following environment :

ATL P7000 robot with FC230 fibrebridge and 4 LTO drives, 3 fibre
connections to switch
Brocade 3800 SAN switch with 16 ports
Netbackup DataCenter 4.5 :
     1 master running on W2K
     8 media servers running on NT and W2K
     Shared Storage option so that all 4 drives can be seen from all 9
media servers
     1 QLogic 2300 in every media server with a connection to switch

Problem :
--------------
When we power cycle the robot for whatever reason, most of the media
servers loose their connection with the drives, i.e. the drives are DOWN on
those servers and we are unable to get them UP. On Windows, it seems that
the drives are available again, but Netbackup is unable to UP them again.
The only way to solve this, is to reboot the Windows server. This is
obviously not wanted as we are talking about production servers.

We already talked with Quantum about this and they told us to use
persistent binding. We are trying to set this up, but are struggling. Can
someone help us with that ? We prefer to find a solution on the robot side,
because when a new media server is installed, we have to go through the
whole process of installing the SAN Surfer Tool Kit and setting up
persistent binding. The SANBlade manager documentation mentions something
about a SANBlade Control FX utility and a wizzard, we don't find that. It
says to start the scfx123.exe program, but we don't find that anywhere
(even the mighty Internet). And how do we know that persistent binding is
active ? The SANBlade tool we are using is not very clear in that.

Who can help us ? Please reply to me individually at
bruno.bossier AT comparex DOT be

Thanks,
Bruno



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