--uuid restore & multiple server pools

Greetings,
Greetings, we have a 3.1.1 environment with 4 disparate server pools (105 servers - 400+ VMs) managed by one centralized Oracle VM Manager host. Each server pool is in a different data center.
We would like to decentralize the management of each server pool (4 total) by standing up a local Oracle VM Manager host within each data center. We use OEM 12c for centralized management and will leverage Zones for role based access.
The goal is to establish a local Oracle VM Manager host in each data center without downtime, i.e. move each pools management from the centralized Oracle VM Manager host to a local Oracle VM Manager in each data center without downtime.
Would a --uuid install at each site allow us to decouple from the centralized Oracle VM Manager host and establish a local Oracle VM Manager host at each data center, i.e. move from one centralized Oracle VM Manager host, to a localized Oracle VM Manager host at each of the 4 sites?
FAQs:
a) currently we have one centralized Oracle VM Manager host managing all four server pools.
b) each pool has localized SRs & networks, i.e. storage and networks are local to each data center.
Thank you for your support!
Edited by: Roddy Rodstein on Jul 31, 2012 9:49 AM

Yes, you can have same servers part of separate server pools, however once server1 in pool1 is assigned then it can't be used again when assigning service profiles. This is done through a "Qualification" process.
Why you might do this? I'm taking this right from the configuration guide..
A server pool contains a set of servers. These servers typically share the same characteristics. Those characteristics can be their location in the chassis, or an attribute such as server type, amount of memory, local storage, type of CPU, or local drive configuration. You can manually assign a server to a server pool, or use server pool policies and server pool policy qualifications to automate the assignment.
If your system implements multi-tenancy through organizations, you can designate one or more server pools to be used by a specific organization. For example, a pool that includes all servers with two CPUs could be assigned to the Marketing organization, while all servers with 64 GB memory could be assigned to the Finance organization.
HTH

Similar Messages

Creating Multiple Server Pools

Hi All,
I have Installed Oracle VM Server 2.2 and Configured HA as well as Installed Oracle Database. Everything is working fine so far.
I have two Partitions from iSCSI and mounted to both the servers(/dev/sdb,/dev/sdc). I have used /dev/sdb to Configure HA and Installed Oracle 11g.
Now I like to Create Second Serverpool on Second VM Server to create test instance on Second Partition(/dev/sdc). Is it possible...?
Will it overwrite the existing /OVS if i crested the Second Serverpool without Cluster since /OVS was mounted to both the servers as part of configure Clustering.
Please could anyone advice me...
Thanks in Advance....

A server can be a member of only one pool. You will have to delete the second server from the first pool, and add it to the new pool. You can select /dev/sdc on that server as the root repository. Since both servers are now in different pools, you will need to disable HA for the first pool.
Edited by: herb on Apr 12, 2010 3:31 PM

Migrating between server pools

Is it possible to move a vm between server pools? Im am trying move a clone vm from one server pool to another server pool...
Im getting the following error: Server Pool: ServerPool has no compatible servers for the clone of ....
Any resaon why i mite be getting that error?
Is it possible to present a repository to multiple server pools?

Exactly how are you trying to do this. Are you choosing to clone and then running the clone customizer to map your storage and vnics?

Server 2012 - Problem with multiple instances of RDS looking to the same license server. Server Pools requiring the wrong servers.

Hi All,
I have two domains, prod.local and test.local. On prod.local, I have a license server living on the connection broker. I have created a two way trust between the domains so that the RDS instance on test.local can use the license server on prod.local.
When I first setup the RDS instance on test.local and point it to use the license server on prod.local everything appears to work, all servers recognize the license server.
However when I log off and back on, and click the RDS dashboard in test.local, the Remote Desktop Services overview states I need to add every server from prod.local to the server pool. This is bothersome because the Remote Desktop Services on test.local
is still functioning, but I am no longer able to manage it. When I add the servers from prod.local to the server pool I am displayed the setup for prod.local instead of test.local
Is there a way around this? Can I use the same RDS license server key multiple times?
tldr: Using a different domain's license server causes my RDS instance in Server Manager to display the wrong RDS farm.

Hi Nathan,
Thank you for posting in Windows Server forum.
Please check the required ports must be opened on firewall.
- To issue RDS Per User CALs to users in other domains, there must be a two-way trust between the domains, and the license server must be a member of the Terminal Server License Servers group in those domains.
- To restrict the issuance of RDS CALs, you can add RDS Host Servers into Terminal Server Computers group on RDS Licensing servers.
- Configure RDS licensing server on all RDS Host Servers in each domain/forest. You can do it through RDS host configuration snap-in or through a group policy.
- Add administrators group of each domain/forest in the local administrators of RDS licensing server. This way, you’ll not get a prompt to enter your credentials when you’ll open RDS host configuration snap-ins in trusted domains/forests.
More information:
Best practices for setting up Remote Desktop Licensing (Terminal Server Licensing) across Active Directory Domains/Forests or Workgroup
http://support.microsoft.com/kb/2473823
Hope it helps!
Thanks.
Dharmesh Solanki

How to start virtual OVM Manager and restore server pool which hosts it?

It had to happen... the day I migrate our OVM Manager from an ESX host into an Oracle VM, we have a massive NetApp outage, and everything fails to come back. As a result of having lock errors when trying to start any VMs I noted on the forum various solutions and ended up following the steps concerning stopping ovs-agent and removal of /etc/ovs-agent/db/* on all the servers. Next I re-added the repo to all the servers using repos.py, but of course now none of the servers are in a pool so the --list command shows them as [R] rather than [*], and the /OVS directory is not yet mapped. So the next stage, according to the instructions which I am following, describes using the 'Restore' option in the OVM Manager GUI. Given that /OVS is not mounted and all of the servers are now config-less, I can't see how I can start the OVM Manager VM to get to the aforementioned "restore" option...
Thank Jeebus I kept the ESX copy of OVM manager on (but powered down), as I was able to fire it up and do the restore, and now everything is hunky dory. However, this is our Dev platform and I was going to build the Prod platform and deploy OVM Manager straight into the OVM architecture via the template, rather than having it on ESX, but this issue on the Dev install has made me think twice about the mobius nature of managing the VM architecture from within the VM architecture...
I am assuming I am missing some sort of commandline knowledge, so... given:
- empty db directory
- root repos [R] on all servers
- as yet unmapped /OVS
...how to get the server pool up and get something back into /OVS, and then start my OVM Manager VM so I can 'restore'? Is it even possible? because if it isn't, I'll be blowed if the Prod OVM manager is going to get deployed like this...
And no, there's no money in the budget for extra tin just to sit there running the OVM Manager and nothing else.
TIA
Kev

user12474530 wrote:
Still nobody seems to know of how to create a server pool on the CLI (outside of OVM Manager) though...There is no way to create a server pool on the CLI without Oracle VM Manager. You either need to use the Web UI of the Manager, or use the ovmcli command-line tool. Both of those require Oracle VM Manager to be running.
However, once the Pool is created, you do not need the Manager to be running to perform simple tasks like starting/stopping guests. That can be done with the xm command-line tool.
I agree with you about running VM Manager as a guest of the Oracle VM Server: I actually don't recommend this in production, unless you have at least two Server Pools. Then you can have the Manager of Pool A hosted by Pool B and vice-versa. Most often however, I recommend a completely independent Manager machine, either physical or hosted on a separate virtualization stack.

Error when trying to create Server Pool

Hi there,
I've been trying to create a server pool in the Oracle VM Manager. However I am getting the following error:
(11/23/2012 12:41:06:471 PM)
OVMAPI_4010E Attempt to send command: create_pool_filesystem to server: gunnar.blpc.com.bb failed. OVMAPI_4004E Server Failed Command: create_pool_filesystem lun /dev/mapper/271480530748fbd8c 2021663b7dffb10a 0004fb00000500004641e478723eecdc 0004fb00000100008126e601b77c2155 0004fb00000200002021663b7dffb10a, Status: org.apache.xmlrpc.XmlRpcException: exceptions.RuntimeError:Command: ['mkfs.ocfs2', '-b', '4K', '-C', '4K', '-J', 'size=4M', '-N', '32', '-L', 'OVS_POOL_FILESYSTEM', '-U', '0004fb00000500004641e478723eecdc', '--global-heartbeat', '--cluster-stack=o2cb', '--cluster-name=2021663b7dffb10a', '--force', '/dev/mapper/271480530748fbd8c'] failed (1): stderr: mkfs.ocfs2 1.8.2
/dev/mapper/271480530748fbd8c is apparently in use by the system; format forced anyway.
stdout:
WARNING!!! OCFS2 uses the UUID to uniquely identify a file system.
Having two OCFS2 file systems with the same UUID could, in the least,
cause erratic behavior, and if unlucky, cause file system damage.
Please choose the UUID with care.
Cluster stack: o2cb
Cluster name: 2021663b7dffb10a
Stack Flags: 0x1
NOTE: Feature extended slot map may be enabled
Proceed (y/N): Aborting operation.
Fri Nov 23 12:41:06 AST 2012
Fri Nov 23 12:41:06 AST 2012
I am connecting to iSCSI storage and I am sure that the oracle vm server is the only device connecting to the target.
Please let me know if anyone has any suggestions.
Thanks

Seems like there's already an OCFS2 fs on that LUN. Could it be, that you tried to create a server pool on that LUN before, or that this target has been used to host some other OVM repository? If that is the case, then you need to wipe the LUN, before you can create a new OVM repository or server pool for that matter on it.
But please… double check that this LUN isn't in use anywhere else, since you could loose data, if you'd wipe it!

How to restore SQL Server 2008 database to SQL Server Express 2012

How to restore SQL Server 2008 database to SQL Server Express 2012.
I BACKED UP a SQL Server 2008 database from work, which runs the enterprise version of 2008, and I tried to RESTORE it to my new 64-Bit home PC which is running the new SQL Server Express 2012 64-Bit (Advanced Edition).
But if fails. I get the following message text.
Can anyone help me please. I thought restoring a 2008 database to the later version 2012 would be possible.
Here is the error message. Thanks.
TITLE: Microsoft SQL Server Management Studio
Restore of database failed. (Microsoft.SqlServer.Management.RelationalEngineTasks)
ADDITIONAL INFORMATION:
System.Data.SqlClient.SqlError: The database was backed up on a server running version 8.00.2039. That version is incompatible with this server, which is running version 11.00.3128. Either restore the database on a server that supports the backup, or use a
backup that is compatible with this server. (Microsoft.SqlServer.SmoExtended)

ADDITIONAL INFORMATION:
System.Data.SqlClient.SqlError: The database was backed up on a server running version 8.00.2039. That version is incompatible with this server, which is running version 11.00.3128. Either restore the database on a server that supports the
backup, or use a backup that is compatible with this server. (Microsoft.SqlServer.SmoExtended)
Are you sure your database is SQL server 2008 from error message it seems your SQL server version is 2000.I guess you have multiple versions of SQL server.
AFAIK you cannot restore backup of SQL server 2000 to 2012 directly .Restore it first to SQL 2008/2008R2 or 2005 then take backup again of new database created and then you would be able to restore it on 2012.Its kind of intermediate way
Also make sure your DB size is less than 10 G as max DB size supported by Express edition is 10 G and if your DB size is more than 10 G restore will fail.
You are also trying to restore enterprise DB backup on express edition so you will loose enterprise features on express edition
Please mark this reply as the answer or vote as helpful, as appropriate, to make it useful for other readers

Oracle vm 3.1.1 server pool file system corruption

Hi
I have a small lab environment composed by 2 Oracle VM Servers and SAN access.
This morning somehow the OCFS2 Server Pool File System (the 12GB one) corrupted and the Server Pool seems dead. How can I recover from this error? I looked in the documentation and MOS but found nothing
Thank you

I know it would be a lot of work in such a scenerio. I don't know what you're using for storage but you could snapshot or replicate the pool storage to create a backup that can be restored.
I do wish they would let you select more than one VM guest in the VM Manager. Would be really nice to have such. Seems like it would be a rather easy feature to implement.

Can not add new VMS into existing clustered server pool

For some reason, we reinstall VMM (3.1.1-416) with existing UUID and rediscover all VMSs. The process went smooth and the whole system looks clean without dead objects any more. The guest VMs are all working fine. However, the new added VMS in unassigned servers are not displayed in the "Available Servers" when we want to add into existing clustered server pool, but it is shown when we add into unclustered server pool. Anyone has clue or possible way to debug what's wrong.
Thanks,
Shun-Jee Liu

The issue is resolved. The access group in shared iSCSI is not properly configured.

What is the Best Practice for Server Pool...

hi,
wht is the Best Practice for having server pool i.e
1) having a single large serverpool consisting of "n" number of guest vm
2) having a multiple small serverpool consisting of less of number of guest vm
please suggest.....

Raja Kondar wrote:
wht is the Best Practice for having server pool i.e
1) having a single large serverpool consisting of "n" number of guest vm
2) having a multiple small serverpool consisting of less of number of guest vm I prefer option 1, as this gives me the greatest amount of resources available. I don't have to worry about resources in smaller pools. It also means there are more resources across the pool for HA purposes. Not sure if this is Official Best Practice, but it is a simpler configuration.
Keep in mind that a server pool should probably have up to 20 servers in it: OCFS2 starts to strain after that.

Running Multiple Server Instance in Weblogic Server

Hi
I am using Weblogic 5.1.0 with SP6. I am trying to run multiple server
instances
of Weblogic Server. The one that comes by default is "myServer".
My question is if i were to create multiple server instances, do i need to
copy the
weblogic.properties file in each of folder corresponding to respective
application
server instance. I am not using Clusters.
Also if i need to include some classes or jar files in my application server
instance
classpath, how do i do it. cuz in weblogic.properties file i didn't find any
property
corresponding to classpath.
The classpath i need to set is for my creating a JDBC pool for connectivity
to my
AS400 DB2 database using the Type 4 (AS400JDBCDriver) Driver. Is there some
other way to do it.
Thanks in advance
Niranjan

There is an answer on your first question at
weblogic.support.interest.clustering
I asked a similar question. Search for a message posted by me on 24/11/00
Rgds,
Mike
"Niranjan Soni" <[email protected]> wrote in message
news:[email protected]..
Hi
I am using Weblogic 5.1.0 with SP6. I am trying to run multiple server
instances
of Weblogic Server. The one that comes by default is "myServer".
My question is if i were to create multiple server instances, do i need to
copy the
weblogic.properties file in each of folder corresponding to respective
application
server instance. I am not using Clusters.
Also if i need to include some classes or jar files in my applicationserver
instance
classpath, how do i do it. cuz in weblogic.properties file i didn't findany
property
corresponding to classpath.
The classpath i need to set is for my creating a JDBC pool forconnectivity
to my
AS400 DB2 database using the Type 4 (AS400JDBCDriver) Driver. Is theresome
other way to do it.
Thanks in advance
Niranjan

JNDI Lookup for multiple server instances with multiple cluster nodes

Hi Experts,
I need help with retreiving log files for multiple server instances with multiple cluster nodes. The system is Netweaver 7.01.
There are 3 server instances all instances with 3 cluster nodes.
There are EJB session beans deployed on them to retreive the log information for each server node.
In the session bean there is a method:
public List getServers() {
List servers = new ArrayList();
ClassLoader saveLoader = Thread.currentThread().getContextClassLoader();
try {
   Properties prop = new Properties();
   prop.setProperty(Context.INITIAL_CONTEXT_FACTORY, "com.sap.engine.services.jndi.InitialContextFactoryImpl");
   prop.put(Context.SECURITY_AUTHENTICATION, "none");
   Thread.currentThread().setContextClassLoader((com.sap.engine.services.adminadapter.interfaces.RemoteAdminInterface.class).getClassLoader());
   InitialContext mInitialContext = new InitialContext(prop);
   RemoteAdminInterface rai = (RemoteAdminInterface) mInitialContext.lookup("adminadapter");
   ClusterAdministrator cadm = rai.getClusterAdministrator();
   ConvenienceEngineAdministrator cea = rai.getConvenienceEngineAdministrator();
   int nodeId[] = cea.getClusterNodeIds();
   int dispatcherId = 0;
   String dispatcherIP = null;
   String p4Port = null;
   for (int i = 0; i < nodeId.length; i++) {
    if (cea.getClusterNodeType(nodeId[i]) != 1)
     continue;
    Properties dispatcherProp = cadm.getNodeInfo(nodeId[i]);
    dispatcherIP = dispatcherProp.getProperty("Host", "localhost");
    p4Port = cea.getServiceProperty(nodeId[i], "p4", "port");
    String[] loc = new String[3];
    loc[0] = dispatcherIP;
    loc[1] = p4Port;
    loc[2] = null;
    servers.add(loc);
   mInitialContext.close();
} catch (NamingException e) {
} catch (RemoteException e) {
} finally {
   Thread.currentThread().setContextClassLoader(saveLoader);
return servers;
and the retreived server information used here in another class:
public void run() {
ReadLogsSession readLogsSession;
int total = servers.size();
for (Iterator iter = servers.iterator(); iter.hasNext();) {
   if (keepAlive) {
    try {
     Thread.sleep(500);
    } catch (InterruptedException e) {
     status = status + e.getMessage();
     System.err.println("LogReader Thread Exception" + e.toString());
     e.printStackTrace();
    String[] serverLocs = (String[]) iter.next();
    searchFilter.setDetails("[" + serverLocs[1] + "]");
    Properties prop = new Properties();
    prop.put(Context.INITIAL_CONTEXT_FACTORY, "com.sap.engine.services.jndi.InitialContextFactoryImpl");
    prop.put(Context.PROVIDER_URL, serverLocs[0] + ":" + serverLocs[1]);
    System.err.println("LogReader run [" + serverLocs[0] + ":" + serverLocs[1] + "]");
    status = " Reading :[" + serverLocs[0] + ":" + serverLocs[1] + "] servers :[" + currentIndex + "/" + total + " ] ";
    prop.put("force_remote", "true");
    prop.put(Context.SECURITY_AUTHENTICATION, "none");
    try {
     Context ctx = new InitialContext(prop);
     Object ob = ctx.lookup("com.xom.sia.ReadLogsSession");
     ReadLogsSessionHome readLogsSessionHome = (ReadLogsSessionHome) PortableRemoteObject.narrow(ob, ReadLogsSessionHome.class);
     status = status + "Found ReadLogsSessionHome ["+readLogsSessionHome+"]";
     readLogsSession = readLogsSessionHome.create();
     if(readLogsSession!=null){
      status = status + " Created ["+readLogsSession+"]";
      List l = readLogsSession.getAuditLogs(searchFilter);
      serverLocs[2] = String.valueOf(l.size());
      status = status + serverLocs[2];
      allRecords.addAll(l);
     }else{
      status = status + " unable to create readLogsSession ";
     ctx.close();
    } catch (NamingException e) {
     status = status + e.getMessage();
     System.err.println(e.getMessage());
     e.printStackTrace();
    } catch (CreateException e) {
     status = status + e.getMessage();
     System.err.println(e.getMessage());
     e.printStackTrace();
    } catch (IOException e) {
     status = status + e.getMessage();
     System.err.println(e.getMessage());
     e.printStackTrace();
    } catch (Exception e) {
     status = status + e.getMessage();
     System.err.println(e.getMessage());
     e.printStackTrace();
   currentIndex++;
jobComplete = true;
The application is working for multiple server instances with a single cluster node but not working for multiple cusltered environment.
Anybody knows what should be changed to handle more cluster nodes?
Thanks,
Gergely

Thanks for the response.
I was afraid that it would be something like that although
was hoping for
something closer to the application pools we use with IIS to
isolate sites
and limit the impact one badly behaving one can have on
another.
mmr
"Ian Skinner" <[email protected]> wrote in message
news:fe5u5v$pue$[email protected]..
> Run CF with one instance. Look at your processes and see
how much memory
> the "JRun" process is using, multiply this by number of
other CF
> instances.
>
> You are most likely going to end up on implementing a
"handful" of
> instances versus "dozens" of instance on all but the
beefiest of servers.
>
> This can be affected by how much memory each instance
uses. An
> application that puts major amounts of data into
persistent scopes such as
> application and|or session will have a larger foot print
then a leaner
> application that does not put much data into memory
and|or leave it there
> for a very long time.
>
> I know the first time we made use of CF in it's
multi-home flavor, we went
> a bit overboard and created way too many. After nearly
bringing a
> moderate server to its knees, we consolidated until we
had three or four
> or so IIRC. A couple dedicated to to each of our largest
and most
> critical applications and a couple general instances
that ran many smaller
> applications each.
>
>
>
>
>

Problem with Server Pool

Hello,
I have an pool of physical servers at the company that supports the production environment of virtual servers.
This pool have 4 (0, 1, 2 and 3) domain servers.
I recently had a problem with the OVS-AGENT, and had to restart the whole pool.
After that, all virtual servers I start at the pool, using the VM Manager, always start at the same domain server, the domain number 3.
When I force the start using xm create, I can start the virtual server at domain server 0, 1 and 2 too.
The live migration only works with the shell command (xm migrate -l), using the VM Manager simple do nothing.
I try to force an especific virtual machine to run at an especific node, using Placement Policy at VM Manager, but when I try to start the server, this error occurs:
Start - /OVS/running_pool/441_n_psgl04
PowerOn Failed : Result - failed:errcode=20000, errmsg=No enough memory to run vm(/OVS/running_pool/441_n_psgl04), required memory=16384.
StackTrace:
File "/opt/ovs-agent-2.3/OVSSiteVM.py", line 124, in start_vm
raise e
Any idea what's happening? Is that some problem with the ovs-agent's database? How can I fix that with minimum of outage of the virtual servers?
I'm really concerned about this problem because this weekend we had a stop on the network that caused the restart of the cluster, and when the servers came back, all virtual servers started up on the same physical server, which caused a lack of resources.
Thanks for your help, and sorry about my poor english!

843474 wrote:
PowerOn Failed : Result - failed:errcode=20000, errmsg=No enough memory to run vm(/OVS/running_pool/441_n_psgl04), required memory=16384.You don't have enough memory on the target server to start this guest. So, you need to work out why you don't have enough memory or why Oracle VM Manager isn't seeing all your servers. Check to see if all the servers are listed as "Active" on the servers tab of Oracle VM Manager. Also, you can try and restore the ovs-agent database to the entire pool by clicking the "Restore" button on the Server Pools tab.

MULTIPLE BUFFER POOL의 개념 및 사용 방법 (ORACLE8)

제품 : ORACLE SERVER
작성날짜 : 1999-05-13
Multiple Buffer Pool의 개념 및 사용 방법
1. 필요성
table이나 index 등 segment는 그 사용 빈도나 중요도 등에 따라 memory에
buffering되는 것을 달리 할 필요가 있다. Oracle8에서는 buffer cache에 대해서
multiple buffer pool이라는 새로운 특성의 개념을 제공하여 segment마다 다른
buffer를 사용할 수 있도록 하고 있다.
multiple buffer pool은 'keep', 'recycle', 그리고 'default' buffer pool로
구성되며, 이것을 control하기 위한 internal algorithm은 하나의 buffer pool을
사용할 때와 대부분 마찬가지다. 즉, 기존의 CACHE option이나 full table scan 시
LRU end에 위치시키는 것 등은 모두 변함이 없으며, 단지 그러한 기법들이 각
buffer마다 별도로 적용된다는 것 뿐이다.
2. buffer의 종류
multiple buffer pool의 주요 목적은 서로 다른 형태로 사용되는 것을 나누어 놓
아 서로 방해가 되지 않도록 하는 것으로 정리할 수 있으며, 각각 다음과 같은
경우에 사용하도록 한다.
(1) KEEP buffer pool : 가능한 한 memory에 오랫동안 유지되어져야 하는
segment를 위해 사용되어져야 한다. 자주 사용되어지고 cache size의 약 10%
전후의 크기를 가진 segment가 이 pool을 사용하기에 적당하다.
그러나, 여기에서도 Oracle7.3의 CACHE option과 마찬가지로 새로이 access
되는 segment에 의해 LRU end 쪽으로 이동하는 것이 가능하므로 항상 cache
된다고 보장할 수는 없다.
적당한 크기로 지정하는 것이 중요한데 당연히, 동시에 memory에 올려지기를
바라는 object들의 크기의 합보다는 커야 한다.
(2) RECYCLE buffer pool : 자주 사용되어지지 않거나, buffer pool의 두배보다
큰 정도의 큰 segment가 index search를 하는 작업 등에 사용되어지도록 한다.
(3) DEFAULT buffer pool : 위의 두 buffer pool에 할당되지 않은 나머지는
default buffer pool이 된다. 그러므로 KEEP이나 RECYCLE buffer pool은
없어도 반드시 default buffer pool은 존재하게 된다.
이 buffer pool은 Oracle7의 하나의 buffer pool과 같다.
3. buffer pool을 설정하는 방법
이러한 종류의 buffer pool을 지정하기 위해서 BUFFER_POOL_KEEP과
BUFFER_POOL_RECYCLE이라는 parameter가 존재하며, DB_BLOCK_BUFFERS
와 DB_BLOCK_LRU_LATCHES parameter도 함께 고려하여야 한다.
syntax는 다음과 같다.
BUFFER_POOL_KEEP=(buffers:<value>,lru_latches:<value>) 혹은
BUFFER_POOL_KEEP=<value>
BUFFER_POOL_RECYCLE=(buffers:<value>,lru_latches:<value>) 혹은
BUFFER_POOL_RECYCLE=<value>
위의 syntax에서 보는 바와 같이 각 pool에 대해서 buffer의 갯수 뿐 아니라
LRU latch의 갯수도 지정할 수 있다. 만약 지정하지 않으면 그 pool에 대해서
하나의 latch가 할당되는 것이다.
DEFAULT pool에 대해서는 명시적으로 block의 갯수나 latch의 갯수를 지정할
수 없고, 대신 전체 block의 갯수 (DB_BLOCK_BUFFERS)와 전체 LRU latch의
갯수 (DB_BLOCK_LRU_LATCHES)에서 KEEP과 RECYCLE에 할당된 각각의
값을 뺀 것만큼 default pool에 할당된다.
간단한 예제로 설명하면 다음과 같다.
예를 들어 initSID.ora file에 다음과 같이 parameter가 설정되어 있다고 가정한다.
DB_BLOCK_BUFFERS=1000
DB_BLOCK_LRU_LATCHES=6
BUFFER_POOL_KEEP=(buffers:400,lru_latches:2)
BUFFER_POOL_RECYCLE=100
이러한 경우 KEEP pool에 대해서는 400개의 block과 2개의 LRU latch가 할당
되고 RECYCLE pool에는 100개의 block과 1개의 LRU latch가 할당된다. 그리
고 DEFAULT pool에는 500 (1000-400-100) 개의 block과 3 (6-2-1)개의 LRU latch
가 할당되게 된다.
각 LRU queue에 대해서 block은 균등하게 배분된다.
즉, 이 예에서 DEFAULT queue는 LRU 1번이 167개의 block을 LRU2도 167개, 그리
고 LRU3은 166개의 block을 가지게 되며, KEEP queue는 두개의 latch가 각각
200개씩의 block을 그리고 RECYCLE queue는 100개의 block을 가지게 된다.
이러한 정보는 v$buffer_pool을 통해 확인이 가능하며,
이 예의 경우 다음과 같이 조회된다. 여기에서 set_count가 각 pool에 할당된
latch의 갯수이며, lo_bnum과 hi_bnum이 buffer의 range이다.
SQL> select * from v$buffer_pool;
NAME      LO_SETID HI_SETID SET_COUNT BUFFERS LO_BNUM HI_BNUM
     0 0 0 0 0 0
KEEP 4 5 2 400 0 399
RECYCLE 6 6 1 100 400 499
DEFAULT 1 3 3      500 500 999
각 queue는 최소 50개의 block은 할당받아야 하며, 그렇지 않은 경우에는 오류
가 발생한다. 즉 예를 들어, BUFFER_POOL_KEEP=(buffers:100, lru_latches:3)과
같이 설정하면 alert.log file에 "Incorrect parameter specification for
BUFFER_POOL_KEEP"이라는 오류 메시지가 적히게 되며, 100개의 block에 대해서
최대 두개의 LRU latch만이 가능하게 된다.
4. buffer pool을 지정하는 방법
BUFFER_POOL이라는 Oracle8에서 새로 추가된 storage 절의 parameter를 이용
하여 segment가 사용할 default pool을 지정할 수 있다. segment의 모든 block은
지정된 pool을 사용하게 되며, 아래의 예제와 같이 사용하면 된다.
CREATE TABLE keep_table(t NUMBER(10)) STORAGE (BUFFER_POOL KEEP);
ALTER TABLE recycle_table storage(BUFFER_POOL RECYCLE);
BUFFER_POOL은 tablespace나 rollback segment에 대해서는 지정할 수 없으며,
clustered table에 대해서는 cluster level에서만 지정이 가능하다. partition
table에 대해서는 각 partition별로 pool을 지정하는 것이 가능하다.
일단 segments가 적당한 pool에 할당이 되고 난 후에는, logical hit ratio나
free buffer waits와 같은 다양한 통계 정보가 확인 가능하다.
이러한 통계 정보를 담고 있는 view는 v$buffer_pool_statistics이며,
이 view는 $ORACLE_HOME/rdbms/admin/catperf.sql을 수행하면 생성된다.

Server Pool WITHOUT shared storage

The documentation defines Server Pool as:
+Logically an autonomous region that contains one or more physical Oracle VM Servers.+
Therefore, should it possible to add multiple servers (physically separate VM servers) to the same Server Pool even though they are NOT using shared storage? When I tried to add the second VM Server to the Server Pool I received the following error:
During adding servers ([vmoracle2]) to server pool (VM_Server_Pool), Cluster setup failed:
(OVM-1011 OVM Manager communication with vmoracle1 for operation HA Setup for Oracle VM
Agent 2.2.0 failed: <Exception: SR '/dev/sda3' not supported: type 'ocfs2.local' not in
['nfs', 'ocfs2.cluster']> )Thanks.

Nothing is as easy as it seems when it comes to Oracle VM.
When trying to create a new Server Pool to accommodate my second VM Server, I received the following error:
+2010-03-08 18:18:24.575 NOTIFICATION Getting agent version for agent:vmoracle2 ...+
+2010-03-08 18:18:24.752 NOTIFICATION The agent version is 2.3-19+
+2010-03-08 18:18:24.755 NOTIFICATION Checking agent vmoracle2 is active or not?+
+2010-03-08 18:18:24.916 NOTIFICATION [Server Pool Management][Server][vmoracle2]:Check agent (vmoracle2) connectivity.+
+2010-03-08 18:18:30.304 NOTIFICATION entering into assign vs action...+
+2010-03-08 18:18:30.311 NOTIFICATION Getting agent version for agent:vmoracle2 ...+
+2010-03-08 18:18:30.482 NOTIFICATION The agent version is 2.3-19+
+2010-03-08 18:18:30.483 NOTIFICATION Checking agent vmoracle2 is active or not?+
+2010-03-08 18:18:30.638 NOTIFICATION [Server Pool Management][Server][vmoracle2]:Check agent (vmoracle2) connectivity.+
+2010-03-08 18:18:45.236 NOTIFICATION Getting agent version for agent:vmoracle2 ...+
+2010-03-08 18:18:45.410 NOTIFICATION The agent version is 2.3-19+
+2010-03-08 18:18:45.434 NOTIFICATION master server is:vmoracle2+
+2010-03-08 18:18:45.435 NOTIFICATION Start to check cluster for server pool+
+2010-03-08 18:18:45.581 WARNING failed:<Exception: Cluster root not found.>+
StackTrace:
File "/opt/ovs-agent-2.3/OVSSiteCluster.py", line 535, in cluster_precheck
clusterprecheck(single_node, ha_enable)+
+File "/opt/ovs-agent-2.3/OVSSiteCluster.py", line 515, in clusterprecheck+
if not cluster_root_sr_uuid: raise Exception("Cluster root not found.")
+2010-03-08 18:18:45.582 NOTIFICATION Failed check cluster for server pool+
+2010-03-08 18:18:45.583 ERROR [Server Pool Management][Server Pool][VMORACLE2_Server_Pool]:Check prerequisites to create server pool (VMORACLE2_Server_Pool) failed: (OVM-1011 OVM Manager communication with vmoracle2 for operation Pre-check cluster root for Server Pool failed:+
+<Exception: Cluster root not found.>+
+)+
+2010-03-08 18:18:45.607 NOTIFICATION Exception Message:OVM-1011 OVM Manager communication with vmoracle2 for operation Pre-check cluster root for Server Pool failed:+
+<Exception: Cluster root not found.>+
The "*Test Connection*" succeeded just fine prior to clicking NEXT on the "Create Server Pool" page.
Any suggestions?

--uuid restore & multiple server pools

Similar Messages

Maybe you are looking for