Im not sure how it works in HACMP, I operate in a sun cluster active/passive
environment and the instances use shares storage and as part of a failover
the disks are 're-homed' to the failover node. When the instance comes
online on the failvoer node it is using all the same files as it was on the
active node, therefore same config.

I would assume (always dangerous) that HACMP uses some similar methodology,
If not could you simply issue the same 'update db cfg' commands on the
passive node so that they are in place should the instance ever be started
on that node?

-----Original Message-----
From: Chuck Hamilton
[mailto:db2udbdba-ezmlmshield-x96668061.[Email address protected]
Sent: Monday, May 24, 2004 3:15 PM
To: LazyDBA Discussion
Subject: Re: Configure Db2/udb for HACMP


Another question for you again regarding HACMP. If I make configuration
parameter changes to the instance and database while it's running on node 1,
what do I need to do to ensure those changes are carried over to node 2 when
the database is activated over there? I guess this questions really boils
down to where are the config parameters stored? In the database, or in a a
file somewhere that I should ensure is synchronized periodically between the
nodes.

I have looked through the Information Center for this stuff but didn't turn
up anything other than "dont modify these files manually".

----- Original Message -----
From: "DANIEL CALLAHAN " <db2udbdba-ezmlmshield-x88223885.[Email address
protected]
To: "LazyDBA Discussion" <[Email address protected]
Sent: Monday, May 24, 2004 2:36 PM
Subject: Re: Configure Db2/udb for HACMP


>
>
>
>
> Did you make sure that the userids and groups are the same numeric on
> each system? ie db2inst1 435 on both systems as well as the groups.
>
> Dan Callahan IT Specialist
> 603-472-4121 t/l 525-4121
> [Email address protected]
> [Email address protected]
>
>
>
>
> "Chuck Hamilton "
> <db2udbdba-ezmlms
> hield-x77055373.x To
> [Email Address Removed] "LazyDBA Discussion"
> a.com> <[Email address protected]
> cc
> 05/24/2004 02:08
> PM Subject
> Configure Db2/udb for HACMP
>
>
>
>
>
>
>
>
>
>
> I have a two node HACMP cluster running active/passive. db2 is
> supposed to fail over to node 2 if node 1 crashes. There is one
> instance and one database which has been running fine on node 1 for
> weeks. When testing failover, db2start works fine on node 2, but
> activating the db fails with SQL5005C. The filesystems have come over
> to node 2 without a problem. To set up node 2 what I've done so far is
> create the same instance owner, das user,
> and fenced user, installed the same version of db2 (8.1), installed the
> same fixpak (5), and cataloged the database at the same directory (which
is
> on one of the filesystems that moves over). What step did I miss?
>
>
>
> ---------------------------------------------------------------------
> PLEASE CLICK REPLY-ALL TO SEND A REPLY TO EVERYONE
> website: http://www.LazyDBA.com
> To unsubscribe, e-mail: db2udbdba-[Email address protected] For
> additional commands, e-mail: db2udbdba-[Email address protected]
>
>
>
>
> ---------------------------------------------------------------------
> PLEASE CLICK REPLY-ALL TO SEND A REPLY TO EVERYONE
> website: http://www.LazyDBA.com
> To unsubscribe, e-mail: db2udbdba-[Email address protected] For
> additional commands, e-mail: db2udbdba-[Email address protected]
>



---------------------------------------------------------------------
PLEASE CLICK REPLY-ALL TO SEND A REPLY TO EVERYONE
website: http://www.LazyDBA.com
To unsubscribe, e-mail: db2udbdba-[Email address protected]
For additional commands, e-mail: db2udbdba-[Email address protected]


DB2 & UDB email list listserv db2-l LazyDBA home page