PI 7.0 Cluster based Installation

Hi all,
         I did Central System installation but never installed High Availability sytem. We got a requirement for a cluster based installation.
I have gone through some links and got a basic idea, but not perfect in starting installation. Could you pls. help me in understanding for a successfully installation. (Requirement: Install SAP Netweaver PI 7.0 in Production Server).
For a High Availability syetm i have seen the following options to follow as per SAPINST:
1.) Central Services Instance for ABAP (ASCS) --   Consists of Msg. service & Enqueue Service (Shared Disk)
2.) Central Services Instance (SCS)                   --   Consists of Msg. service & Enqueue Service (Should be installed after installation of 1.)                                                                                (Same Shared Disk)
3.) Microsoft Cluster Service (MSCS)                 --   This step creates the SAP cluster group on the first MSCS node and adds the ASCS                                                                                and SCS instance(s) to the SAP cluster group
4.) Database Instance
5.) Microsoft Cluster Service (MSCS)
6.) Enqueue Replication Server
7.) Central Instance                                           -- If you install the central instance on one of the MSCS nodes, make sure that you use a                                                                                local drive, not a shared drive.
8.) Dialog Instance                                            -- You have to install at least one dialog instance. Make sure that you install the dialog                                                                                instance on a host other than the central instance host.
>> I guess that we need to install ASCS(1) and SCS(2) in two seperate systems, but what is that shared disk?(Aslo mentioned that NFS must be installed?) [It means that we require 2 hosts?]
>> Mentioned that one Enqueue Replication Server (ERS) for each Java SCS and one ERS for each ABAP SCS (ASCS) installed in your system. [It means that we require 2 hosts?]
>> Do we need to install all the above mentioned instances only in Windows or will this work on any platform?? [Example: SOLARIS]
>> Database Instance on 1 host, Central instance in 1 host --- That means we require a total of 6 hosts (excluding Dialog Instance??)
Pls. suggest whether my undertanding is correct or not???
Thanks & Regards,
Subbu

Dimitrios,
Check the below links
Installation
High Availability
Cheers
Agasthuri

Similar Messages

Moving Physical SQL Cluster into Virtual SQL Cluster based on Hyper-V Failover

Hello All............I have a SQL Cluster based on Physical Hardware that has Three-3 instances as well. I have setup a Hyper-V Failover Cluster (2012 R2) and have built Virtual/Guest SQL Cluster (2012 R2) upon it. Now, I intend to move
the instances/databases from Physical SQL Cluster to Virtual SQL Cluster.
1. Is this supported? If so, I would appreciate any guidance on it?
2. Is P2V of SQL Cluster supported in Hyper-V Failover Cluster based on Windows Server 2012 R2?

Hi Sir,
Please refer to the following blog regarding moving SCCVM SQL database to another SQL server:
http://blogs.technet.com/b/configurationmgr/archive/2013/04/02/how-to-move-the-configmgr-2012-site-database-to-a-new-sql-server.aspx
It is quoted from the similar thread :
https://social.technet.microsoft.com/Forums/en-US/a1558842-cdf5-4e5f-8f10-d660e96eae1b/migration-sql-for-sccm-2012?forum=configmanagermigration
But it seems that is a migration of system center production , I would suggest you to post the question to system center forum ：
https://social.technet.microsoft.com/Forums/en-US/home?forum=configmanagermigration%2Coperationsmanagergeneral&filter=alltypes&sort=lastpostdesc
Best Regards,
Elton Ji
We
are trying to better understand customer views on social support experience, so your participation in this
interview project would be greatly appreciated if you have time.
Thanks for helping make community forums a great place.

Software Updated-Based Installation Conundrum

To my understanding, software update-based installation of the Config Manager client agent works by importing it into WSUS (which is a requisite component for the server with the software update point role) where it is classified as a critical
update. So far so good.
However, while a WSUS server can automatically approve an update's installation, the actual download and installation of an update is dependent upon the configuration of a client's Windows Update
policy. This local policy needs to be configured to use the SUP server as its WSUS server, and it likely must have some kind of schedule to download and install the update (unless someone is taking upon themselves the unenviable task of manually
updating computers).
Thus, it would seem to make good sense to use an Active Directory group policy to ensure that these WSUS settings are configured correctly on computers when they join a company's domain. Unfortunately, in my environment we've found that such a
group policy creates a conflict, because once a computer has the Config Manager client, SCCM will then require uncontested control over the Windows Updates policy. The resultant conflict causes update deployments to fail with a "group policy
conflict" error and, worse still, client computers will wind up running two separate SVCHOST processes that result in uncontrolled memory consumption.
I've contacted Microsoft software assurance support on this issue, and they found it no less a catch-22 than I have. I'd be interested to hear how others handle SU-based installation effectively in their environment. I suppose in many places, it's simply
effected on workstation images. In others, perhaps the settings are applied directly as registry edits in group policy preferences. I'd be especially interested to find out if anyone has created a WMI filter than can distinguish clients with the agent from
those without it.

Some of your understanding above is sort of correct. Ultimately, WSUS is *only* used within ConfigMgr to download the update catalog from Microsoft and make it available to clients. That's it. The Windows Update Agent (WUA) is still used, however, it does
so by direction of the ConfigMgr agent which means that nearly all of the settings in group policy for Windows Update are meaningless in a ConfigMgr environment.
Jason, thanks for the response. It seems you're impugning my understanding while simultaneously reiterating my explanation. Either I was unclear, or I'm not grocking the distinction you're trying to draw. Let's see if we can get on the same
page.
To summarize, SU-based installation results in the client being treated as an approved update on the WSUS component, and then it's downloaded and installed according to the schedule defined in the client's Windows Update policy. Once the client
is installed, Config Manager takes over and updates are centrally managed through deployments issued to collections.
However, perhaps where we're having a disconnect is that, going forward, computers which don't have the agent yet still require Windows Update to be configured for them to ensure they download and install it. New computers come into the
environment perpetually. Applying a group policy in AD would seem an ideal solution, but in practice such a policy has created the aforementioned conflict on computers that do have the agent.
I have actually read your blog posts during the course of researching this issue, as well as many others. They largely treat WSUS configuration via AD GPO's as undesirable in concert with SU-based installation (for reasons that you outline) but
they don't concern themselves with how administrators can be assured that new arrivals consistently download and install the client agent via SU-installation method.

Change Master Agent for Cluster-Based Agent -- OEM 11.1.0.1

        We had a 6 node RAC cluster and a cluster-based management agent installed to monitor.
        Node 6 was the Master agent.
       Now node 6 is not part of the cluster.
    I am getting metric collection for the cluster.
I cluster configuration in OEM shows Master agent as node 6. I only have 3 nodes now.
How do I update the master agent or the cluster.
Thanks,
Silver

On the clustef homepage in OEM there is a link called properties under the name of the cluster. Click the link it will shlw the true master node. What I have seen in the pasf what can rsolve the issue is bounce the agent on that master node. You can also try a clear state. You can also use the configuration for the cluster database to make sure the old instance has been remove from OEM.

Web-based installer

I'm looking to create a Web-based installer program, but have absolutely no clue how it can be implemented.
Would you guys shed me some light on this :)

I think WebStart is what you're looking for (although we may just be talking over each other here). WebStart downloads a Java program to the client via a server but doesn't necessarily have to be a client/server type app. It's pretty slick since it can be configured to check for new version of the program and update it as needed. Try installing WebStart and playing around with some of the programs yourself.

OIM9.1.0.1 in Cluster mode installation issue

Hello,
I am facing some issues to setup OIM9.1.0.1 with JBoss Application Server (Cluster mode). After installation of OIM, when I try to start application server, then it is not getting stated. Below are the some initial JBoss AS errors
2008-12-18 12:14:59,625 ERROR [org.apache.commons.digester.Digester] Begin event threw exception
java.lang.ClassNotFoundException: No ClassLoaders found for: org.jboss.web.tomcat.tc5.sso.ClusteredSingleSignOn
     at org.jboss.mx.loading.LoadMgr3.beginLoadTask(LoadMgr3.java:212)
     at org.jboss.mx.loading.RepositoryClassLoader.loadClassImpl(RepositoryClassLoader.java:521)
2008-12-18 12:15:00,656 ERROR [org.jboss.deployment.MainDeployer] Could not start deployment: file:/C:/thcOIMSetup/jboss-4.2.3.GA/server/all/deploy/jboss-web.deployer/ROOT.war/
org.jboss.deployment.DeploymentException: URL file:/C:/thcOIMSetup/jboss-4.2.3.GA/server/all/deploy/jboss-web.deployer/ROOT.war/ deployment failed at org.jboss.web.tomcat.service.TomcatDeployer.performDeployInternal(TomcatDeployer.java:386)
2008-12-18 12:15:00,703 ERROR [org.jboss.deployment.MainDeployer] DeploymentException while trying to deploy a package with a new deployer
org.jboss.deployment.DeploymentException: URL file:/C:/thcOIMSetup/jboss-4.2.3.GA/server/all/deploy/jboss-web.deployer/ROOT.war/ deployment failed
The environment details are
•     OIM 9.1.0.1
•     jdk1.6.0
•     JBoss-4.2.3.GA
•     Apache Web Server 2.2
•     Oracle11g
I am referring OIM 9.1.0 documentation; “Deploying in a Clustered JBoss Application Server Configuration“. I am not able to figure out the root cause of the problem. Please provide me some pointer to handle this problem.
Any help would be greatly appreciated!
Regards,
Hardew

Hi,
I was able to resolve this issue, it looked like the main Tomcat deployer wasn't working i.e. the xlWebApp wasn't getting loaded, infact none of the webapps were getting loaded, i checked it out and found the classloader was trying to load the classes which jboss 4.2.3GA version doesn't even recognize at first place and hence the root.war deployment was failing and which triggered the failure for all the other deployment.
I changed the classloader to load the classes which were right for jboss 4.2.3GA, for e.g.
"org.jboss.web.tomcat.tc5.sso.ClusteredSingleSignOn" to "org.jboss.web.tomcat.service.sso.ClusteredSingleSignOn"
"org.jboss.web.tomcat.tc5.jca.CachedConnectionValve" to "org.jboss.web.tomcat.service.jca.CachedConnectionValve"
after changing this the main tomcat deployer worked and deployed Root.war and after that everything else got deployed.
You can change this by changing these values in the jboss-web.deployer/server.xml or do the custom classloading using a separate custom classloader if you don't want to touch the jboss 4.2.3GA setup after OIM 9101 install.
Thanks
Shivam
Edited by: Blue Divided on Jan 9, 2009 1:05 PM

Problem in Domain Server for Cluster SAP installation

Hello,
SAP Production Landscape Information :
SAP Version: ECC 6.0
OS: Windows 2003
Database : Oracle 10.2.0.2
Production Server is in MSCS cluster involving two nodes.
Also, this is a domain installation and hence user SAPSID and SAPSERVICESID resides on altogether different domain Server.
Problem:
OS of u2018Windows domain serveru2019 has crashed. Hence, domain server is almost not usable.
Restart of it takes several hours. Repair of OS is tried by Microsoft without any success.
So this has resulted into communication problem between SAP Production server and Domain server. As an effect, restart of production server takes approximately an hour.
And hence I need some solution to resolve this SAP Server  Domain Server link.
Logically, following can be solution of this problem,
1.     If Current SAP Production can be made as Local Installation instead of Domain installation. (Basically, creating users locally and making SAP refer that)
2.     If above is not feasible then replicating domain server on new hardware and making SAP point to this new domain server.
However, I am not sure about feasibility and procedures for both of above cases.
Please let me know your views and any other solution for the same.
Thanks and Regards,
Nirav.

Hi Markus,
Thanks for your reply.
Possibility of converting domain installation to local is ruled out as MSCS cluster needs domain.
As you have rightly mentioned, I do have secondary domain sever.
For More information:
1)Domain details for primary server are
    Full computer name : APPL-SERVER.SHREE
    domain            : SHREE
   as secondary server is kept shutdown, details are not available. (I can't run both server running simultaneously)
2)sqlplus "/as sysdba"
sql> select username from dba_users;
    TSMSYS
    OPS$SHREE\SAPSERVICEPRD
    OUTLN
    OPS$SHREE\PRDADM
    DIP
    DBSNMP
    SAPPRD
    SYS
    SYSTEM
3) In the services, mainly users are using SHREE\PRDADM & PASSWORD.
Client have tried shutting down primary and promoting secondary domain controller. However, after that SAP service were not getting started. But on again starting primary controller,start of SAP services were successful.
As can be seen in 2nd & 3rd point above,SAP database and SAP services both refer to name "SHREE" which seems domain name and not to hostname.
So in this case, when I start secondary domain server (primary is shut down) then as SAP is referring to domain name, SAP's service should start.
However, that was not the case. SAP Services were not able to get started.
So, do we have to change anything in SAP as to make it point to secondary domain server.
Let me know your views.
Regards,
Nirav.

Hide a View Variant embedded inside View Cluster based on Switch Assignment

Hi,
We have a view cluster in which one 2 view variants are present.
Now, we want to display only one of these view variants based on the switch assignment.
While creating an entry for view variant in View cluster I've added swithc name to one of the views. So based on whether the switch is active or not that view variant will be displayed.
But for the other view variant there is no switch assignment made, so it will always be visible.
Is there any way that we can make the visiblity of 2 view variants mutually exclusive based on some switch assignment
Any help is appreciated.
Regards
Manas Dua

closing the thread..

Import a column from an excel spreadsheet, then enable/disable boolean controls in a cluster based on imported values

Greetings all. I have an application I've been developing and refining for some time. Part of it entails an interactive periodic table of elements where the user can select multiple elements to include in a multi-element chemical standard. The part I'm having difficulty with now is disabling multiple elements within the periodic table. Specifically, what the scientist wants is to be able to reference an excel inventory that they keep of available on-hand single-element standards. I need to be able to read in column A of that spreadsheet, build a distinct array of those elements listed (some are listed more than once) and then disable the boolean controls in my periodic table for those elements not in the array (or enable those elements which are in the array, whichever makes themost sense). This way, the analyst cannot select elements in the periodic table for which thay have no on-hand inventory. I see how to enable or disable all controls in a boolean cluster, and I see how to enable/disable a single control, but I'm looking for an elegant way to import the excel column, build an array of thsoe values and then enable only those values. Any help is always appreciated. Thanks again.

hi
try this
Gaurav k
CLD Certified !!!!!
Do not forget to Mark solution and to give Kudo if problem is solved.
Attachments:
Excel column read.vi ‏29 KB

SQL Cluster 2014 installation error the following is the right solution ? if yes why ? if not what are the pros n const of this ?

http://www.bradg.co.za/?p=12
SQL 2008 R2 Install Permissions
SQL 2008 R2 is a little tricky to install if you don’t have the right permissions. If the account you are using does not have the correct permissions you will find yourself getting loads of strange errors such as below during the installation.
not all privileges or groups referenced are assigned to the caller
The process does not possess the ‘SeSecurityPrivilege’ privilege which is required for this operation
The error will prompt you to retry or cancel. If you cancel your SQL installation will fail and if you retry you will be prompted with the error again.
The way to resolve this issue is to grant the account you are using the correct privileges.
Local Administrator on the box you are installing SQL
Goto > Local Security Policy > User Rights Management
Back up files and directories
Debug Programs
Manage auditing and Security log
Restore files and directories
Take ownership of files or other objects
Add your installation user account to the above User Rights Management
Once this has taken affect you should be able to install SQL without any issues.

I might render OK in browsers, but the code is riddled with errors.
http://validator.w3.org/check?verbose=1&uri=http%3A%2F%2Fwww.incredibleart.org%2Flinks%2Fa rtists_black.html
This one looks like it could be your red herring.
Line 111, Column 83: cannot generate system identifier for general entity "f.defaultView.getComputedStyle"
Nancy O.

Solaris 8 Patch Cluster 200801 installation / disk utilization spikes

Hello All,
Since we installed the latest Patch Cluster patch our disk utilization has shot up substantially. Is there anyone having same experience? If you do, could you please help us to identify some areas to take a look?
Thanks very much.
Yi

hi again. i just fixed my problem. apparently when i unzipped the file and then burned it onto cd...the data got corrupted. i fixed it by copying the zip file onto cd and then extracting the patches onto the solaris machine.

Processor based installations........

during the installation of oracle apps on a multi processor
box, how does one ensure that oracle will utilize all
available processors ????
in another scenario, we have oracle apps on a dual
processor box. we have imminent plans of adding
two more cpu's to this box. what changes should
we make to the oracle apps installation so
that the two extra cpu's will be used by oracle ???
we have oracle8i with apps11i on solaris2.8 on
a sun4500.
thanks & regards

Subbiah,
Use the SUN Java 1.4.1 Plugin instead of JInitiator
Frank

Web-based installation was unsuccessful

Yesterday, when I attempted to start the AnyConnect Secure Mobility Client to establish a long-working VPN connection, the software failed to automatically install the client (see attached screen capture). I tried clicking on the Window 7 link to manually install. The manual install seemed to work but the client still did not connect. Any idea what I should do to once again be able to establish a VPN connection? I am running 3.1.04063 of the client

No, it's hardly a broblem of anyconnect image. Most probably it's got smth to do with java or smth like that on your client pc. Check prerequisites for web-installation of anyconnect (i mean browser settings, java settings and version, etc).

Suggestion for UEFI based installs

Dear all,
Going through the Beginners install guide, I realized that the partitioning section mentions that for UEFI based systems, a UEFI partition needs to be installed.
However, there's something I'd like to share. I hope it can be changed if required. I'd be happy to help.
1) The sequence of making the UEFI partition. Should it be made first (beginning), followed by other system partitions?
2) The UEFI section says that "boot" flag needs to be set for this partition, which is fine. However, the Install guide says that the root partition needs to be set "bootable". Is there any potential conflict?
3) When mounting the system partitions, the /boot/efi directory needs to be mounted to the UEFI system partition in order to install GRUB. This is mentioned in the GRUB UEFI install guide, not in the Beginners' Guide.
Last edited by hdhiman (2012-08-29 05:25:04)

1) It is common for it to be on the beginning of the drive, but it doesn't really matter. Boot flag is doing all the magic.
2) Root shouldn't be flaged bootable. Anyways, there is no way you are booting from anything on the drive but the EFI partition.
3) It should be there and you can add it yourself.

Cluster's Installations

Hi Everyone,
Today I uninstalled my CRS and reinstalled the CRS, during the re-installation I got this error OUI-25031 Oracle Notification server configuration assitant failed. Can you please suggest me how to overcome this error. Thanks in advance for your help.

OS-Red Hat Enterprise Linux-4
CRS-11g
Failed- At when OUI is trying to install the Oracle Notification Server, I think this is the last step.
Actually I am trying this setup on a Vmware, I uninstalled the CRS from OUI, created new disks and mapped those raw devices for the OCR, Voting Disks and ASM for the fresh installation next time, even then I encountered the same error.

PI 7.0 Cluster based Installation

Similar Messages

Maybe you are looking for