SEO

#1 (**permalink**) 01-16-2008, 06:13 PM

i have 2 servers, QAS and PRD. They are on serviceguard. QAS is the
adoptive node and PRD is the primary node. I am using the one package
concept. is my shutdown procedure fine?

1) execute first for production
#cmviewcl -> returns the package name
#cmhaltpkg –v mypackage (as root user)
#cmviewcl -> check that the package is stopped
#cmhaltcl -> halt the cluster
#cmviewcl -> checked that the cluster is not running
Shutdown the Production UNIX System, execute:
#shutdown –hy 0

2) do the same for qas
#cmviewcl -> check that the cluster has been shut down
#stopsap

are the above proedures fine?

PS : the server startup scripts contains commands to start the cluster
package and all services relating to SAP are also started when cluster
package is up and running.

#2 (**permalink**) 01-16-2008, 06:14 PM

In article <c06e4d68.0407150600.67a2c1f6@posting.google.com >, yls177 wrote:
>
>
> i have 2 servers, QAS and PRD. They are on serviceguard. QAS is the
> adoptive node and PRD is the primary node. I am using the one package
> concept. is my shutdown procedure fine?
>
> 1) execute first for production
> #cmviewcl -> returns the package name
> #cmhaltpkg –v mypackage (as root user)
> #cmviewcl -> check that the package is stopped
> #cmhaltcl -> halt the cluster
> #cmviewcl -> checked that the cluster is not running
> Shutdown the Production UNIX System, execute:
> #shutdown –hy 0
>
> 2) do the same for qas
> #cmviewcl -> check that the cluster has been shut down
> #stopsap
>
> are the above proedures fine?
>
Hey;

Why do you need to do a stopsap on the adoptive node after the
package and cluster is already down? Isn't it dead as a result
of the package going down?

Other than that, it'll certainly work...

Doug

--
--------
Senior UNIX Admin
O'Leary Computer Enterprises
dkoleary@olearycomputers.com (w) 630-904-6098 (c) 630-248-2749
resume: http://www.olearycomputers.com/resume.html

#3 (**permalink**) 01-16-2008, 06:14 PM

> i have 2 servers, QAS and PRD. They are on serviceguard. QAS is the
> adoptive node and PRD is the primary node. I am using the one package
> concept. is my shutdown procedure fine?
>
> 1) execute first for production
> #cmviewcl -> returns the package name
> #cmhaltpkg -v mypackage (as root user)
> #cmviewcl -> check that the package is stopped
> #cmhaltcl -> halt the cluster
> #cmviewcl -> checked that the cluster is not running
> Shutdown the Production UNIX System, execute:
> #shutdown -hy 0

> 2) do the same for qas
> #cmviewcl -> check that the cluster has been shut down
> #stopsap
>
> are the above proedures fine?

It seems correct at first glance. I would do the same thing. Your
approach is more robust than just shutting down both nodes (which
would also work, assuming you disable package switching first).
Stopping the package(s) and the cluster manually insures that
everything is stopped gracefully.

Question: why do you run stopsap on QAS? Do you have another non-HA
instance running on it?

--
Olivier S. Masse

#4 (**permalink**) 01-16-2008, 06:14 PM

omasse@iname.com (Olivier S. Masse) wrote in message news:<4daa0929.0407151734.4be367bb@posting.google. com>...
> > i have 2 servers, QAS and PRD. They are on serviceguard. QAS is the
> > adoptive node and PRD is the primary node. I am using the one package
> > concept. is my shutdown procedure fine?
> >
> > 1) execute first for production
> > #cmviewcl -> returns the package name
> > #cmhaltpkg -v mypackage (as root user)
> > #cmviewcl -> check that the package is stopped
> > #cmhaltcl -> halt the cluster
> > #cmviewcl -> checked that the cluster is not running
> > Shutdown the Production UNIX System, execute:
> > #shutdown -hy 0
>
> > 2) do the same for qas
> > #cmviewcl -> check that the cluster has been shut down
> > #stopsap
> >
> > are the above proedures fine?
>
> It seems correct at first glance. I would do the same thing. Your
> approach is more robust than just shutting down both nodes (which
> would also work, assuming you disable package switching first).
> Stopping the package(s) and the cluster manually insures that
> everything is stopped gracefully.
>
> Question: why do you run stopsap on QAS? Do you have another non-HA
> instance running on it?

thanks guys... i am pretty new to serviceguard. i dont think i have
another non-HA instance running on my QAS.

this is confirmed by doing a "cmviewcl -v"
i get to see my cluster, 2 nodes (one in the hostname of PRD, and the
other in the hostname of QAS) and the package having the status UP

and also a ps -ef | grep dw only shows one sap sid.

i was told that using the below to shutdown was not appropriate as
compared to what i have post. any comments

(for both prd and qas, hostname)
cmhaltnode hostname
cmrunnode hostname

#5 (**permalink**) 01-16-2008, 06:14 PM

> i was told that using the below to shutdown was not appropriate as
> compared to what i have post. any comments
>
> (for both prd and qas, hostname)
> cmhaltnode hostname
> cmrunnode hostname

No, this will not do anything useful. It will move your package(s) to
QAS (if package switching is active) and still leave the node inside
the cluster.

SeviceGuard commands and their output isn't very intuitive. I suggest
you install ServiceGuard Manager (available on software.hp.com), it's
a GUI for ServiceGuard that will help you do most of your operations.

Olivier S. Masse

#6 (**permalink**) 01-16-2008, 06:14 PM

omasse@iname.com (Olivier S. Masse) wrote in message news:<4daa0929.0407180724.2c52235f@posting.google. com>...
> > i was told that using the below to shutdown was not appropriate as
> > compared to what i have post. any comments
> >
> > (for both prd and qas, hostname)
> > cmhaltnode hostname
> > cmrunnode hostname
>
> No, this will not do anything useful. It will move your package(s) to
> QAS (if package switching is active) and still leave the node inside
> the cluster.
>
> SeviceGuard commands and their output isn't very intuitive. I suggest
> you install ServiceGuard Manager (available on software.hp.com), it's
> a GUI for ServiceGuard that will help you do most of your operations.
>
> Olivier S. Masse

are u referring to sgmgr

#7 (**permalink**) 01-16-2008, 06:15 PM

> > SeviceGuard commands and their output isn't very intuitive. I suggest
> > you install ServiceGuard Manager (available on software.hp.com), it's
> > a GUI for ServiceGuard that will help you do most of your operations.
> >
> > Olivier S. Masse
>
> are u referring to sgmgr

Yes Sir!

Olivier S. Masse

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

SEO

shutdown procedures for sap on serviceguard