Re: Configure failover in seconds

You can modify the isAlive and looksAlive timeouts, as well as the number of times to restart a resource, and the interval between restarts via cluadmin. Right click your disk resource, go to properties, then advanced.

Adjust the looksAlive and isAlive timeouts to suit your needs. You'll probably have to experiment to find the optimal settings since you don't want to ignore failed resources.

Good luck!

~kenny

"Edwin vMierlo [MVP]" <EdwinvMierlo@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message news:e7rtpCdrHHA.1168@xxxxxxxxxxxxxxxxxxxxxxx
Ok, here is the real answer: you can not.

When a failover starts, it starts. There is no parameter where you can set a
"pause" because of your san disks who need to failover.

What you can do however is to create a "scripted geo-cluster"

1) script your "failover commands"
2) add this script as a resource to your cluster
3) make your disk resources dependent on your script

this way, on failover, your script fires on node 2 first, your disks are not
failed over in the san / still at node 1, but your script takes care of
this, and your disk resources do not come online until your script finishes.

script finish / online --> disks on san are failed over --> physical disk
resources come online properly

and please see this blog (
http://msmvps.com/blogs/jtoner/archive/2007/05.aspx ) from John Toner on
this subject, it details how you do this on an EMC Clariion, but you
probably can apply methods to your specific environment.

Please realize, if you do all this, you are NOT using a qualified solution,
I am afraid you are running a NON-supported solution.

I would never run production applications on such a scripted geo-cluster.
Because I can think of many, many scenarios where your script would not
execute or not pick up the exact conditions to determine the correct action.
Also a scripted geo-cluster will have a risk of a split brain due to not
picking up the correct information, or not having the logic to deal with the
situation/failure.

Testing with this in a test/lab environment, sure ! I have done it, good fun
!

rgds,
Edwin.



"Mike" <Mike@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:80F5ABDB-B401-4256-B998-1501A6ABA929@xxxxxxxxxxxxxxxx
Guys, all I want to "please" know is where do you set the number of
seconds
to wait before failover occurs from one node to another in Windows Server
2003? Is this done in CLuster Administrator? If so, where do I set this
property?

Thank you.



--
Mike


"Edwin vMierlo [MVP]" wrote:

> If you are running a geographically dispersed cluster, what SAN /
software
> are you using for this ?
>
>
> "Mike" <Mike@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
> news:C9B6F40E-474D-4EB6-AA9C-6AEC335D2C62@xxxxxxxxxxxxxxxx
> > I want to change the number of seconds MSCS failover occurs so it > > does
not
> > conflict with SAN failover.
> >
> > Thank you.
> >
> > -- > > Mike
> >
> >
> > "Rodney R. Fournier [MVP]" wrote:
> >
> > > What is your business need? I understand the technical requirement.
> > >
> > > Cheers,
> > >
> > > Rodney R. Fournier
> > >
> > > MVP - Windows Server - Clustering
> > > http://www.nw-america.com - Clustering Website
> > > http://msmvps.com/clustering - Blog
> > > http://www.clusterhelp.com - Cluster Training
> > > ClusterHelp.com is a Microsoft Certified Gold Partner
> > >
> > >
> > > "Mike" <Mike@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
> > > news:5843AC3C-6B38-4F5F-BF2A-0C22623AD5B4@xxxxxxxxxxxxxxxx
> > > >I want to set a property in Cluster Administrator: number of
seconds to
> > > >wait
> > > > before failover occurs from one node to another in Windows Server
> 2003.
> > > >
> > > > Thank you.
> > > >
> > > > -- > > > > Mike
> > > >
> > > >
> > > > "Rodney R. Fournier [MVP]" wrote:
> > > >
> > > >> http://support.microsoft.com/kb/815267
> > > >>
> > > >>
> > > >> To set the value of the ClusSvcHeartbeatTimeout property, run > > > >> the
> > > >> following
> > > >> command from a command prompt:
> > > >>
> > > >> cluster.exe /cluster:clustername /prop
clussvcheartbeattimeout=number
> of
> > > >> seconds
> > > >>
> > > >> where clustername is the name of the cluster and number of
seconds is
> the
> > > >> number of seconds that you want to use in the calculation of the
> > > >> heartbeat.
> > > >>
> > > >> But, I would not mess with it. What are you trying to > > > >> accomplish.
> > > >>
> > > >> Cheers,
> > > >>
> > > >> Rodney R. Fournier
> > > >>
> > > >> MVP - Windows Server - Clustering
> > > >> http://www.nw-america.com - Clustering Website
> > > >> http://msmvps.com/clustering - Blog
> > > >> http://www.clusterhelp.com - Cluster Training
> > > >> ClusterHelp.com is a Microsoft Certified Gold Partner
> > > >>
> > > >>
> > > >> "Mike" <Mike@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
> > > >> news:87BC79B2-217D-4995-92CF-3C2497996173@xxxxxxxxxxxxxxxx
> > > >> > Where do you set the number of seconds to wait before failover
> occurs
> > > >> > from
> > > >> > one node to another in Windows Server 2003?
> > > >> >
> > > >> > Thank you.
> > > >> > -- > > > >> > Mike
> > > >>
> > > >>
> > > >>
> > >
> > >
> > >
>
>
>



.

Relevant Pages

  • Are sysctl(8) values useful for measuring system resource consumption?
    ... the "developer experience," what resources are thus consumed, and figure ... script that essentailly acts as a bit of "scaffolding" around time; ... the script sets things up to invoke timewith the "-l" flag (so we ...
    (freebsd-performance)
  • Re: Trusted sites issue...
    ... Have you ensured that your script does not include references ... If the script uses resources ... from an unsecured area, or sends to a different site, it may be ... Internet Explorer Security Options, Part 1 ...
    (microsoft.public.windows.inetexplorer.ie6.browser)
  • Re: How to Get WebResources.axd working on IIS on XP Pro
    ... with whatever script resources your objects need. ... Rick Strahl shows you how to embed your own resources for use with WebResources.axd: ... I have an aspx script that works fine when run from Visual Web ... I think I've tracked the problem to a failure of IIS to find ...
    (microsoft.public.dotnet.framework.aspnet)
  • Re: outlook clean free/busy script?
    ... I'm fairly certain that the script would only act on the "primary" calendar ... > I am using this program to manage my resources: ... > Users are using all versions of Outlook. ... If I gave userA permission ...
    (microsoft.public.exchange.admin)
  • Re: Cluster could not fail over
    ... cables, you shouldn't expect the node to failover its resources. ... > So in this case is the Heartbeat the cause of the problem? ...
    (microsoft.public.windows.server.clustering)