Re: cluster scripting suggestions?
- From: "ahill74@xxxxxxxxx" <ahill74@xxxxxxxxx>
- Date: 22 Mar 2007 11:19:57 -0700
Gotcha. You're probably correct to say it is non-HCL Chuck. The MS
recommendation was to use SQL log shipping as our method of data
synchronizing data between the 2 sites. Attempts have been made at
doing just that, but it was an error prone process, and often required
intervention.
I actually believe I've figured out a way to do what I want, for
anyone interested. It'll need some more fine tuning, but here's the
guts of what'll be in our batch file:
cluster ClusterName group "ClusterGroup" /prop | findstr
"PersistentState *1 (0x1)$" > C:\path\cluster_group_status.txt
type nul>0
fc C:\path\cluster_group_status.txt 0
if not %errorlevel% == 0 goto online
:offline
REM SAN replication can begin
goto beginreplication
:online
REM The cluster group is online, next line will take offline
cluster ClusterName group "ClusterGroup" /off /wait:120
goto beginreplication
:end
:beginreplication
REM Add normal processes to initiate replication
Basically, if PersistentState 1(0x1) isn't found in the results of
the /prop query, the file size of the text file created will be 0
bytes, and it'll determine that the cluster is not online.
On Mar 22, 11:53 am, "Chuck Timon [MSFT]"
<cti...@xxxxxxxxxxxxxxxxxxxx> wrote:
http://www.microsoft.com/windowsserver2003/techinfo/overview/san.mspx
Chuck Timon, Jr.
Microsoft Corporation
Longhorn Readiness Team
This posting is provided "AS IS" with no warranties, and confers no rights.
<ahil...@xxxxxxxxx> wrote in message
news:1174569118.172442.87170@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Thanks for the comments guys.
I don't believe that we'd really invest too much time into re-thinking
our replication strategy, and it's impact on our live systems at the
backup site. Our Sharepoint system is just a small piece of our
environment, and it's the only system that will have their resources
online. All other clustered systems will be up, but the cluster
resources won't be online unless we actually have some type of
catastrophe....or periodic testing.
Chuck,
I'm not certain what you meant by non-HCL, pardon my ignorance. But, I
suppose this could be considered a geo-cluster, as the replication
does traverse our internal network, via a gigabit pipe across town.
The Sharepoint environment on both sides are mirrors of one another,
with independent clusters. If the primary system fails, we'll just
alter DNS to point to the backup site.
tks.
On Mar 22, 7:26 am, "Chuck Timon [MSFT]" <cti...@xxxxxxxxxxxxxxxxxxxx>
wrote:
So, basically this is a Non-HCL GeoCluster?
Chuck Timon, Jr.
Microsoft Corporation
Longhorn Readiness Team
This posting is provided "AS IS" with no warranties, and confers no
rights.
<ahil...@xxxxxxxxx> wrote in message
news:1174497836.762155.73740@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Hello all,
I was hoping someone may have some suggestions for my scenario. My
company has 2 SAN environments, one in production, and another in our
disaster recovery site. The servers are identical at both ends. At a
specific time of day, data replicates from the production SAN to the
disastery recover site SAN. Our MS Sharepoint system uses a 2 node
cluster to house all of its data. We want to avoid any data corruption
by ensuring the cluster disk resources in the disastery recovery site
are offline when data replication begins.
Basically, what I'd like to achieve is to script some minor cluster
logic before we kick off replication. This would all be written into a
batch file, or .vbs. I've read through all the command line
documentation, and haven't found out all that I need to do, and
haven't had any luck googling the pants out of it either.
Here's the basic goal, a Scheduled task runs batch file which does the
following:
1. Is cluster group "x" on cluster "xyz" online?
2. If yes, take cluster group "x" offline. Begin data replication.
3. If no, begin replication.
I'm guessing I can achieve the same goal by omitting the 'If yes/no
then do this' portion, but I think it would probably be best practice
to verify that the cluster resource is indeed online before attempting
to take it offline.
I've determined that I can check cluster status:
C:\>cluster ClusterName group "ClusterGroupName" /stat
or cluster properties
C:\>cluster ClusterName group "ClusterGroupName" /prop
and look for a 0 or 1 associated with the "PersistentState" attribute
I've also determined how to take the desired resources offline via the
command line. What I don't know is how to pipe the output of the /stat
or /prop switches into a usable variable, and how to use that variable
to determine the next step for the batch file to take.
Any suggestions? Or RTFM resources to point me to?
Thanks,
Ajh
.
- References:
- cluster scripting suggestions?
- From: ahill74@xxxxxxxxx
- Re: cluster scripting suggestions?
- From: Chuck Timon [MSFT]
- Re: cluster scripting suggestions?
- From: ahill74@xxxxxxxxx
- Re: cluster scripting suggestions?
- From: Chuck Timon [MSFT]
- cluster scripting suggestions?
- Prev by Date: Re: MOM Workgroup Edition and MS Clusters
- Next by Date: Re: MSDTC conflicting articles
- Previous by thread: Re: cluster scripting suggestions?
- Next by thread: Cluster stops responding
- Index(es):
Relevant Pages
|
