Veritas Cluster Server on Solaris x64

Hi All,
We're trying to configure VCS to failover SCS. However, when we bring SCS up via VCS it starts the instance, but VCS doesn't seem to reconize that the instance is up. Therefore after a bit, VCS thinks the instance is not up and tries to failover to the second node.
We have seen that there is a SAP agent for VCS on Solaris sparc, but can't find one for x64.
Does anybody know of an agent for x64, or has anybody configured a standard resource type for SCS in VCS?
Any help will be greatly apreciated.

Wow!
Well luckily we managed to work something out with the stardard "App" resource type in VCS. We monitor whether our SCS instance is up by monitoring the following processes:
en.sap<SID>_SCSxx pf=/usr/sap/<SID>/SYS/profile/<SID>_SCSxx_vhostname
ms.sap<SID>_SCSxx pf=/usr/sap/<SID>/SYS/profile/<SID>_SCSxx_vhostname
<SID> = System ID
xx = SCS instance number
vhostname = the hostname of the virtual ip used for the cluster group.
Hope this helps somebody in the future.

Similar Messages

How to install & confgiure Oracle9i on Veritas Cluster Server

Hi,
I want to know how to install oracle9i on Veritas Cluster Server having OS as Solaris 9. Also How to configure oracle in VCS so that it can be failover.
I really appreciate the replies.
Thanks,
Shrikant

I want to install following component
Oracle 9i form server & report server
Oracle HTTP Server in SSL mode
Oracle As Web Cache
Regards
Arvind

JES & Veritas Cluster Server

Hi All,
By this message i want to share with you my experience with Veritas and JES.
Last week, i configured the Veritas Cluster Server Agent for iPlanet 5.2 to work with JES 6.1. There was problem to bring the ressource group of JES online because the agent is't conpatible with the JES 6, but done somes modification to allow the agent to monitor my JES ressource group.
Firstly :
When bringing the JES ressource group onlin, the VCS Agent search for a file named /etc/msgregistry.inf, this file deasn't exist in JES 6 only in 5.2. to avoid this problem i created an empty file in both nodes using touch :
touch /etc/msgregistry.inf
Secondly :
The old binaries directory sbin on the messaging server was located on the msg-bin and in the new version JES 6 it's located in sbin and to avoid this problem i created a symbolic link like :
ln �s serverroot/sbin serverroot/msg-sbin
Where the serverroot is the full path name to the messaging directory configured on the Veritas Cluster.
I finded this issues only by debugging the Agent scripts, probably you will be faced to other problem, i advice you to just look at the veritas log and try to circumvent the problem.
Enjoy my help.
Mourad.

I don't feel like you've read my response.
It appears you're using the 5.2 cluster agent. You need the new JES cluster agent. It's on the JES CD.
There is no longer an "inf" registry file for it to look at. There is documentation for installing cluster. Please examine and follow.
http://docs.sun.com/app/docs/doc/819-2650/6n4u4dtn1?a=view

Sunray Server Sizing - Solaris x64

Hi,
Where can I obtain information on how to size a Sunray server running on x4170 hardware for CPU and memory in a pure kiosk configuration where the Sun connector redirects the session to a XP VM? Currently I base my config on 20 - 35 users per CPU core and 256MB ram per user. Also with network configuration as in how do I team NIC's together using Solaris 10 for the above configuration, is it needed or what are the best practices.
Thank you.

On the SunRay wiki are some sizing examples. Memory is the crucial part, they say.

Solaris Cluster Server Homogeneous System Copy

Hi,
We have SAP ECC 6.0 on Solaris Sparc server. Our database is Oracle 10.2.0.2. We will buy 2 new solaris server and we want to use these server as cluster. We want to make homogeneous system copy from current server to new cluster server. How can we make homogeneous copy from current server to new cluster server? Firts of all do we have to install new ECC 6.0 on cluster server and then will we make homogeneous system copy?
Also how can we install ECC 6.0 on Solaris Cluster server? Will we have install ECC 6.0 seperatly for each server of cluster unit? I have SAP installation document but it is not clear. Do you have another document for SAP installation on cluster server? Please help us about these issues.
Best regards.

First to move your ECC system from your current server to the new server, yes you do need to homogenous copy. There are other methods like storage sub-system level copy, but it is not supported. Since your system is ABAP, you can setup standby database on the new server and use dataguard to replicate database. You can manual move profiles, binaries and other application file systems manually since they are on external storage. You should change your profiles to use logical hostname instead of the physical hostname though Solaris 10 gives you the ability to move zone/containers with same hostname.
I am assuming you have distributed system, ie database and SAP separate. You can setup RAC for database HA. For SAP, you can use sun cluster or veritas cluster to move SAP central instance with the logical hostname to the 2nd standby server.
You can also separate ASCS from SAP central instance and setup failover for the CS instance with the cluster software. Check notes 821904, 870652. You can take another step in HA with ENQ replication (see http://help.sap.com/saphelp_nw70/helpdata/EN/36/67973c3f5aff39e10000000a114084/content.htm)
-Regards

Veritas Cluster 6 + Solaris 11 + Oracle RAC 11g2 = OCR trouble

I trying new version of Veritas Cluster and Solaris.
Experienced trouble with clustered VxFS for OCR file.
Grid installer refuse VxFS with message - not support the storage type.
After installation I tryed to add new OCR file on VxFS got message in crsd.log:
==============
+2013-01-14 11:05:38.898: [ OCROSD][26]utstoragetypecommon: Oracle Cluster Registry does not support the storage type configured. OCR can be configured on: ASM, NFS, Character Device, VxFS+
+2013-01-14 11:05:38.898: [ OCROSD][26]utdvch:-1: New location /app/oracle/ocrvote2/2.ocr configured is not valid storage type. Return code [37].+
+2013-01-14 11:05:38.898: [ OCRRAW][26]propriodvch: Error [8] returned device check for [app/oracle/ocrvote2/2.ocr]+
+2013-01-14 11:05:38.898: [ OCRRAW][26]dev_replace: master could not verify the new disk (8)+
File system mounted on both nodes with option mincache=direct.
What can be the reason for this error?

Trouble was solved by VCS patch 6.0.3
Version 6.0.1 does not support Solaris 11.1

Question about Grid Control Agents on a Veritas Cluster (or any cluster)

Hi, I'm seeking advice on how to setup Grid Control agents on a Veritas Cluster.
My cluster is made up of two Sun M4000 servers running Solaris 10 x64 (node1 and node2), and 8 virtual hosts setup to fail between the nodes. Each of these virtual hosts has several mount points from a SAN, and contain Oracle 9 and 10 homes as well as databases. The versions are 9.2.0.8 and 10.2.0.3.
I have a full grid control agent install on each node to mount points that don't fail over to the other nodes. These agents monitor the nodes themselves.
Now for my question, is it better to do full intalls of the grid agent in each virtual host or to use the "emctl deploy agent" from the full installs already on node1 and node2 to the virtual hosts? Or is there some other, better way to do this.
Either method results in 9 grid control agents on a node if all the virtual hosts are on that box. I've experimented with both agent deploys and installs a bit and found that the cpu usage goes though the roof with that many agents on one server. Maybe I don't have them configured correctly, but every now and then they spike cpu usage to 100%, and these servers have 8, dual core processors in them with 32GB of RAM.
Thanks for any tips!
Currently, my setup looks something like this, if it helps:
ORANODE1...........................................ORANODE2
user: oracle..........................................user: oracle
full agent install....................................full agent install
VirtualHost1
user: ora_user1
Oracle 10G Home
repository database
OMS install
full agent install
VirtualHost2
user: ora_user2
Oracle 9i Home
production database
full agent install
VirtualHost3
user: ora_user3
Oracle 9i Home
production database
full agent install
VirtualHost4
user: ora_user4
Oracle10g Home
production database
full agent install
VirtualHost5
user: ora_user5
Oracle10g Home
production database
full agent install
VirtualHost6
user ora_user6
Oracle10G Home
production database
full agent install
VirtualHost7
user: ora_user7
Oracle10G Home
production database
full agent install
VirtualHost8
User: ora_user8
Oracle10G Home
production database
full agent install

Will the agent 10.2.0.3 work with Grid Control 10.2.0.1
Yes.

Veritas Cluster (SFRAC) and oracle CRS

We have oracle 9i RAC running on Solaris 9 with over veritas cluster.
We want to run 10g RAC on the same same servers which will use the
veritas cluster file system
While trying to install 10g CRS the node selection list doesnt show node
information.Entered information using response file .After the installation
root.sh fails to start CRS.
Is it ok to run 9i and 10g RAC on same servers.

Chandra/Vishwa
We have a libskgxn2.so in /opt/ORCLcluster/lib directory. But the size of the file is
different from /opt/VRTSvcs/rac/lib/libskgxn2_64.so.1 file
Since we already have 9i rac running, I am not sure if copying the file from
/opt/VRTSvcs/rac/lib directory if it will affect 9i rac.
I went ahead with installation of CRS. I am getting following error while running root.sh
Checking to see if Oracle CRS stack is already configured
Checking to see if any 9i GSD is up
/oracle/crshome/AUTEST/bin/lsdb: Failed to initialize Cluster Context
skgxn error number 1311719766
operation skgxnqtsz
location SKGXN not av
errno 0: Error 0
Setting the permissions on OCR backup directory
Setting up NS directories
Oracle Cluster Registry configuration upgraded successfully
clscfg -install -nn nodeA,nodeAnum,nodeB,nodeBnum... -o crshome
-l languageid -c clustername -q votedisk
[-t p1,p2,p3,p4] [-pn privA,privAnum,privB,privBnum...]
[-hn hostA,hostAnum,hostB,hostBnum...]
With CRS and veritas running on the same server , will CRS always communicate
across nodes using VCS since we will be using veritas cluster file system
for database.
Shankar

Sun Cluster Server restore

Recently one of cluster server crashed without backup. I had to reinstall all of Solaris 8 and software such as veritas volume manager and oracle, Sun Cluster 3.0. Unfortunately, it is said that if one of both cluster servers was down, I have to reinstall both machines. Is it true?
Is there anyone can give me some advices how to fix this problem properly?
Thxs

I think I am having same problem as you before. If possible, could you please give me some instructions. I am just a beginner of Sun Cluster Server. According to your sugguestion, it looks so complex that I can handle. Presently, I have installed volumn manager and oracle on the broken node. Can you instruct me what's next? The more details the better.
Thanks in advance.

Veritas volume manager on Solaris 10(X86)

is it possible to install veritas volume manager on solaris 10 (x86) ?

You should check the supported operating system matrix for Storage Foundation to see what they support. But yes, Solaris x64 is supported. I don't think they support 32-bit x86.
Darren

Installing SOA Suite 11gR1 + OSB 11gR1 on Solaris x64

Experts,
I'm having some trouble installing SOA Suite and OSB on Solaris x64.
According to the last certification matrix
[http://www.oracle.com/technetwork/middleware/downloads/fmw-11gr1certmatrix.xls]
ALL products are supported on Solaris x64 (version 11.1.1.4+) , but when trying to execute the runInstaller I get the following error:
$ ./runInstaller -jreLoc /usr/jdk/instances/jdk1.6.0/
./runInstaller: /soasoat/install/SOA/Disk1/install/solaris/runInstaller: cannot execute
I searched in eDelivery but there aren't any SOA installers (for version 11.1.1.1) for Solaris on x64
Is there any way to bypass this error?
Thanks in advance.
Leandro

Hi Leandro
1. SOA Installer has nothing to do with 32/64 bit. I mean there is NO specific 64 bit version of soa software. It is the underlying Weblogic Server that matters.
2. Since you have 64 bit machine, very first thing, you should install a 64 Bit JDK (Sun or JRockit). And make sure to put this in path. When you run java -version, you should see this 64 bit jdk referred. Now comes Weblogic Server. You need to Download Weblogic Server Generic Installer (file ending with .jar). Then you install Weblogic server using command something like this:
java -d64 -jar weblogic_generic_xxxx.jar (Try -d or -D). After installation, now you have Weblogic Server 64 Bit installed on top of 64 Bit JDK.
3. Now run the SOA Installer and for the jre, you can refer the above jdk/jre location.
Thanks
Ravi Jegga

How to update a server from Solaris 10 MU6 to Solaris 10 MU8

Hi guys,
I've a problem here that it should be easy to solve for some GURU experts like you :-)
I've a server with Solaris 10 10/08 (Update 6) and I need to go to Solaris 10 10/09 (Update 8).
Looking at Release note and other relevant Sun documentation doesn't provide a clear method to upgrade from MU6 to MU8.
Three things are important here:
1) I need to preserve data and applications, so this MUST BE an upgrade, not a FRESH INSTALL
2) we don't have a CD/DVD in the server.
3) we have LDOMs installed on this server (the Guest Domain and the Logical Domains need to be patched).
Can anybody tell me if this is possible and if there is any good document describing it?
P.S. - unless the solution provided into the article "http://blogs.sun.com/patch/entry/solaris_10_10_08_patch" is the only one, I would prefer an update and not a patch.
Thanks in advance
Sergio

My guess is that you have already completed this upgrade, but in the event that you haven't I recommend that you use NFS to either mount a DVD from another server or else copy the OS DVD to a local drive on a server that has a DVD drive and then NFS mount the file system on the server you are trying to upgrade. I use NFS fairly often for upgrades and for patching. I have one server that has a file system that contains the source code from the installation DVD and another file system that contains the most current patch cluster.

Rapid clone 11.5.10 from single node to Veritas cluster environment

Is anybody there who had done the same and working? Please shed some light.
version 11.5.10
Source
DB tier : DB, CM and Admin ( OS : Solaris 10 ), single node
App tier : FrmSrv, iAS ( OS : Solaris 10), single node
Target
DB tier : DB, CM, and Admin ( OS : Solaris 10), 2-nodes Veritas Cluster
(active/passive)
App tier : FrmSrv, iAS (OS : Solaris 10), 2-nodes load balancer
I've many time single node to single node rapid clone , no problem.
But with this single node to multi-nodes, no luck.
your input is much appreciated

Hi
You can use the Advanced cloning (section 4) of the cloining document# 230672.1
to clone signle node to multi node systems.
for webserver load balancing follow the metalink note#:217368.1
Regards
Srinath

Veritas Cluster failover for oracle

Hi,
I'm performing veritas cluster failover for 2 oracle servers which is not working in one way. I've 2 solaris boxes where veritas cluster is configured for failover. In Box1 I'm running 10g and Box2 9.2 when I failover from Box1 to Box2 is working but the other way is not working I'm using 10g listerner for both the services.
Any help would be appreicated.
Thanks,

Hi,
I'm performing veritas cluster failover for 2 oracle servers which is not working in one way. I've 2 solaris boxes where veritas cluster is configured for failover. In Box1 I'm running 10g and Box2 9.2 when I failover from Box1 to Box2 is working but the other way is not working I'm using 10g listerner for both the services.
Any help would be appreicated.
Thanks,

Veritas Cluster

Hi
I am using Veritas Cluster 4.1 in Solaris 9 enviroment.
I have two nodes system and I need to shutdown these two nodes together for maintenance.
I should run hastop -all and then shutdown?
How to start up these two nodes (together or in some sequence?)
best regards

There is no command in Veritas cluster to simultaneously start up
both cluster nodes, you could write a script to run the hastart on one node them rsh to the other node and run hastart if you had to .

Veritas Cluster Server on Solaris x64

Similar Messages

Maybe you are looking for