EJB lookup in a cluster (DNS names to servers mapping)

          I am using weblogic6.0 to setup a object clusters. Currently I have 2 physical
          machine each running one object server.
          I have a servlet in the web tier (it is not part of the cluster) that lookups
          the EJB's.
          From the documentation, I understand intialContext lookup should use one DNS name
          that maps to both the object servers.
          My question is how can I setup that ?
          - I tried the following, I am not sure if this is right
          I created a virtualhost with
          name: MyVirtualHost
          virtual host names : MyCluster (this is name of the cluster I had setup )
          Targets : MyCluster
          I tried looking up with
          t3//MyVirtualHost:7001 and t3//MyCluster:7001
          No luck in both the cases !!!!!
          Note : I restarted my weblogic servers after
          creating the virtual host. But I did not reboot my machine ( don't know if this
          is required )
          Please help !!!!
          Thanks
          Abi -

          You need add an entry in Domain Name Server to include your host1 and host2, and
          use it in your JNDI lookup. Or you can do:
          t3://host1,host2:7001
          Both ways should work. Will load balancing to two JNDI trees.
          Jim Zhou.
          "Abinesh Puthenpurackal" <[email protected]> wrote:
          >
          >I am using weblogic6.0 to setup a object clusters. Currently I have 2
          >physical
          >machine each running one object server.
          >
          >I have a servlet in the web tier (it is not part of the cluster) that
          >lookups
          >the EJB's.
          >
          >From the documentation, I understand intialContext lookup should use
          >one DNS name
          >that maps to both the object servers.
          >
          >My question is how can I setup that ?
          > - I tried the following, I am not sure if this is right
          > I created a virtualhost with
          > name: MyVirtualHost
          > virtual host names : MyCluster (this is name of the cluster I had
          >setup )
          > Targets : MyCluster
          >
          >I tried looking up with
          >t3//MyVirtualHost:7001 and t3//MyCluster:7001
          >
          >No luck in both the cases !!!!!
          >
          >Note : I restarted my weblogic servers after
          >creating the virtual host. But I did not reboot my machine ( don't know
          >if this
          >is required )
          >
          >Please help !!!!
          >
          >Thanks
          >Abi -
          >

Similar Messages

JNDI lookup in a cluster

Hi,
          The WL documentation contains the following:
          "When clients obtain an initial JNDI context by supplying the cluster DNS
          name, weblogic.jndi.WLInitialContextFactory obtains the list of all
          addresses that are mapped to the DNS name"
          Should clients (e.g. an EJB) be concerned about retrying the call to get the
          Initial Context. In other words, can this call fail if one of the servers in
          the cluster fails?
          After the intital contect is obtained, it seems that the lookup should
          always work (since WL will take care of individual server failures and retry
          the lookup in needed).
          Not clear whether the call to get the initial context is guaranteed to
          succeed (as long as one server in the cluster is up, of course)... Any
          information would be appreciated.
          Thanks,
          Philippe


          Hello Philippe,
          I had posted a similar question but now can't find it...got lost I suppose. Anyway,
          I wanted to add in my findings on this. I have a stateful session object running
          in a clustered setup. This stateful object has Home references to multiple stateless
          beans. When I create a failover my stateful object does its failover properly.
          But, if I don't perform a new Home lookup for the stateless objects needed I receive
          the following error:
          ####<Nov 9, 2001 2:00:06 PM CST> <Error> <> <gwiz> <testServer1> <ExecuteThread:
          '9' for queue: 'default'> <> <> <000000> <<TestDeliveryActionHandler>Problem occured
          when trying to do a save and goto. java.rmi.NoSuchObjectException: Activation
          failed with: java.rmi.NoSuchObjectException: Unable to locate EJBHome: 'GBTestManagerHome'
          on server: 't3://10.1.17.3:7001
          When I perform a lookup during the ejbactivate() method to get a new Home reference
          all seems to work OK. My question though is, is this correct? From what I have
          read I had the same impression that the unserialized Home reference should be
          able to locate a new reference in the cluster without having to perform a lookup
          again.
          Any advice from anyone is greatly appreciated,
          Rich
          "Philippe Fajeau" <[email protected]> wrote:
          >Hi,
          >
          >The WL documentation contains the following:
          >
          >"When clients obtain an initial JNDI context by supplying the cluster
          >DNS
          >name, weblogic.jndi.WLInitialContextFactory obtains the list of all
          >addresses that are mapped to the DNS name"
          >
          >Should clients (e.g. an EJB) be concerned about retrying the call to
          >get the
          >Initial Context. In other words, can this call fail if one of the servers
          >in
          >the cluster fails?
          >
          >After the intital contect is obtained, it seems that the lookup should
          >always work (since WL will take care of individual server failures and
          >retry
          >the lookup in needed).
          >
          >Not clear whether the call to get the initial context is guaranteed to
          >succeed (as long as one server in the cluster is up, of course)... Any
          >information would be appreciated.
          >
          >Thanks,
          >
          >Philippe
          >
          >

WLS6.0 Cluster DNS

          I've got only broken pieces of information about setting up a cluster DNS name
          (so that you can use it as url for JNDI connection) from WLS6.0 documentations.
          One place it says (actually suggested) to use cluster's DNS name, the other place
          says that the DNS name for the cluster can be defined by the weblogic.cluster.name
          property (Didn't say how ?) and it is a command-line property which is set when
          the cluster is started. I'm sure it is not the "Name" of a cluster in the Admin
          console. May be it is set up in the host file?
          Thanks,
          Lieyong


Hi,
          Yes you could set up your hosts file as
          myCluster 192.168.1.11
          myCluster 192.168.1.12
          myCluster 192.168.1.13
          and use "t3://myCluster:7001" to lookup your clustered EJB.
          Peace,
          Lynch
          "Lieyong Fu" <[email protected]> ¼¶¼g©ó¶l¥ó news:3b8e7dcf$[email protected]..
          >
          > I've got only broken pieces of information about setting up a cluster DNS
          name
          > (so that you can use it as url for JNDI connection) from WLS6.0
          documentations.
          > One place it says (actually suggested) to use cluster's DNS name, the
          other place
          > says that the DNS name for the cluster can be defined by the
          weblogic.cluster.name
          > property (Didn't say how ?) and it is a command-line property which is set
          when
          > the cluster is started. I'm sure it is not the "Name" of a cluster in the
          Admin
          > console. May be it is set up in the host file?
          > Thanks,
          > Lieyong

DNS name lookup in clustering environment

          Hi,
          I have 2 managed servers hosted on different IP addresse in a cluster.
          I have configured Cluster address in weblogic console to these 2 IP addreses seperated
          by a comma.
          If I am doing lookup for a session bean from my application, which URL i should
          use in lookup?
          If I have a DNS name configured, how I should use it? Can I lookup using DNS name
          only?
          For Ex this is what I have done-
          I have 2 managed servers : 10.1.30.155 and 10.1.30.151
          I have configured cluster addresses to '10.1.30.155, 10.1.30.151'
          I am using url- 't3://DNSName' for Session Bean lookup.
          Can any one guide me on this?
          Thanks in advance.
          Prasanna


Hi,
For the manual create the CNO usually cause may issue, you'd better don't use that method.
The related article:
Windows 2008/ Windows 2008 R2 CNO Recovery
http://blogs.technet.com/b/hugofe/archive/2010/03/11/windows-2008-windows-2008-r2-cno-recovery.aspx
Hope this helps.
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.

Cluster network name resource 'Cluster Name' failed registration of one or more associated DNS name(s) for the following reason: The handle is invalid.

I'm stuck here trying to figure this error out.
2003 domain, 2012 hyper v core 3 nodes. (I have two of these hyper V groups, hvclust2012 is the problem group, hvclust2008 is okay)
In Failover Cluster Manager I see these errors, "Cluster network name resource 'Cluster Name' failed registration of one or more associated DNS name(s) for the following reason: The handle is invalid."
I restarted the host node that was listed in having the error then another node starts showing the errors.
I tried to follow this site: http://blog.subvertallmedia.com/2012/12/06/repairing-a-failover-cluster-in-windows-server-2012-live-migration-fails-dns-cluster-name-errors/
Then this error shows up when doing the repair: there was an error repairing the active directory object for 'Cluster Name'
I looked at our domain controller and noticed I don't have access to local users and groups. I can access our other hvclust2008 (both clusters are same version 2012).
<image here>
I came upon this thread: http://social.technet.microsoft.com/Forums/en-US/85fc2ad5-b0c0-41f0-900e-df1db8625445/windows-2012-cluster-resource-name-fails-dns-registration-evt-1196?forum=winserverClustering
Now, I'm stuck on adding a managed service account (mas). I'm not sure if I'm way off track to fix this. Any advice? Thanks in advance!
<image here>

Thanks Elton,
I restarted 3 hosts after applying the hotfix. Then I did the steps below and got stuck on step 5. That is when I get the error (image above). There
was an error repairing the active directory object for 'Cluster Name'. For more data, see 'Information Details'.
To reset the password on the affected name resource, perform the following steps:
From Failover Cluster Manager, locate the name resource.
Right-click on the resource, and click Properties.
On the Policies tab, select If resource fails, do not restart, and then click OK.
Right-click on the resource, click More Actions, and then click Simulate Failure.
When the name resource shows "Failed," right-click on the resource, click More Actions, and then click Repair.
After the name resource is online, right-click on the resource, and then click Properties.
On the Policies tab, select If resource fails, attempt restart on current node, and then click OK.
Thanks

EJB lookup question

Suppose I have a cluster of two servers with DNS name server1 and
          server2. From an external java program, if I do a JNDI lookup using the
          host name of "server1" and look up an EJB that exists on server1, I
          should always get a reference to an EJB on server1, right? If I use a
          host name of "server1", I should never get connected to server2's naming
          server, right?
          Kirk


Kirk Everett <[email protected]> wrote in message news:<[email protected]>...
          > Suppose I have a cluster of two servers with DNS name server1 and
          > server2. From an external java program, if I do a JNDI lookup using the
          > host name of "server1" and look up an EJB that exists on server1, I
          > should always get a reference to an EJB on server1, right? If I use a
          > host name of "server1", I should never get connected to server2's naming
          > server, right?
          >
          > Kirk
          Kirk,
          not sure about the naming, but you could get an EJB from either
          server. To stop this you have to use a specific property
          (PIN_TO_PRIMARY_SERVER) which we haven't got to work yet!
          Matt

EJB lookup failing in DCM managed Oracle AS clustering

Hi
My purpose is to test DCM managed Oracle AS clustering:
For that I have two Oracle Application servers installed in two different machines with
Version: 10.1.2.0.2 and Installation Type as J2EE and Web Cache
I created an OracleAS File-based Farm and associated both the instances of AS to the cluster. Then I created One OC4J instance called InstanceOne on the cluster and deployed my Application ear onto the cluster.I also configued web Application and EJB Application State Replication the with OracleAS Cluster (OC4J).
My Application has one Ejb module and a web module. My stateless Session bean is Spring framework implemented and I am trying to load the Spring Context lookup from the Struts Action Class by using WebApplicationContext.
My spring context looks like this
<bean id="jndiTemplate"
class="org.springframework.jndi.JndiTemplate">
<property name="environment">
<props>
<prop key="java.naming.factory.initial">com.evermind.server.rmi.RMIInitialContextFactory</prop>
<prop key="java.naming.provider.url">opmn:ormi://<server host>:InstanceOne/<Application Name></prop>
<prop key="java.naming.security.principal">oc4jadmin</prop>
<prop key="java.naming.security.credentials">admin</prop>
<prop key="oracle.j2ee.naming.cache.timeout">30</prop>
<prop key="Dedicated.Connection">yes</prop>
<prop key="Dedicated.RMIcontext">false</prop>
<prop key="LoadBalanceOnLookup">true</prop>
</props>
</property>
</bean>
<bean id="service"
class="org.springframework.ejb.access.SimpleRemoteStatelessSessionProxyFactoryBean">
<property name="jndiTemplate">
<ref bean="jndiTemplate"/>
</property>
<property name="jndiName">
<value>MyServiceEJB</value>
</property>
<property name="cacheHome">
<value>false</value>
</property>
<property name="refreshHomeOnConnectFailure">
<value>true</value>
</property>
<property name="businessInterface">
<value>com.my.springejb.MyBusinessInterface</value>
</property>
</bean>
When I give only one server lookup like
opmn:ormi://<server host1>:InstanceOne/<Application Name>
then lookup is working, But by giving two servers in cluster, like opmn:ormi://<server host1>:InstanceOne/<Application Name>, opmn:ormi://<server host2>:InstanceOne/<Application Name>
I am getting error like this
06/12/01 19:27:41 java.lang.NumberFormatException: For input string: "InstanceOne"
06/12/01 19:27:41 at java.lang.NumberFormatException.forInputString(NumberFormatException.java:48)
06/12/01 19:27:41 at java.lang.Integer.parseInt(Integer.java:468)
06/12/01 19:27:41 at java.lang.Integer.parseInt(Integer.java:518)
06/12/01 19:27:41 at com.evermind.server.rmi.RMIInitialContextFactory.getInitialContext(RMIInitialContextFactory.java:237)
06/12/01 19:27:41 at javax.naming.spi.NamingManager.getInitialContext(NamingManager.java:662)
06/12/01 19:27:41 at javax.naming.InitialContext.getDefaultInitCtx(InitialContext.java:243)
06/12/01 19:27:41 at javax.naming.InitialContext.init(InitialContext.java:219)
06/12/01 19:27:41 at javax.naming.InitialContext.<init>(InitialContext.java:195)
06/12/01 19:27:41 at org.springframework.jndi.JndiTemplate.createInitialContext(JndiTemplate.java:105)
06/12/01 19:27:41 at org.springframework.jndi.JndiTemplate.execute(JndiTemplate.java:83)
06/12/01 19:27:41 at org.springframework.ejb.access.SimpleRemoteStatelessSessionProxyFactoryBean.create(SimpleRemoteStatelessSessionProxyFactoryBean.java:19)
My intension of testing is, if I make one server instance (InstanceOne) down then my lookup should divert all ejb calls to the other server in the cluster, which is not happening at all, and I am getting Server SHUT DOWN exception in the server.
Can anybody please let me know, whether the approach is correct and is there any problem with Spring lookup which does not support multiple host lookup or anything I am missing in my steps, as I am following the Oracle Application Server High Availability Guide 10g Release 2 (10.1.2)

you can find an running sample from :
http://javahowto.blogspot.com/2007/12/calling-weblogic-ejb-3-from-jruby-and.html
In weblogic, local ejb don't appear in the jndi tree, but , it can be founded ...
It works into 10.0 and 10.3
Cheers !

Unable to lookup EJBHOme in Cluster!!

          HELP!
          I am running a client program that attempts to call a stateless EJB (MediaAgent) on a cluster called
          mycluster running on port 63889. The cluster contains objects (EJBs) only.
          Here's the wierd part. When the client performs a listing of the EJBs on mycluster, it shows
          the bean I want, MediaAgentHome.
          But when it tries to do a lookup, I get this error:
          The client was unable to lookup the EJBHome
          The deployment of thje EJB with JNDI name MediaToolHome on the WebLogic Server t3://mycluster:63889
          What in God's name is going on here?!?!
          It can list it but can't look it up??
          Any ideas?
          Thanks,
          Dave


Hi,
          You should use t3://<managed-server1-ip>:<managed-server1-port>,<managed-server2-ip>:<managed-server2-port> as your cluster's provider-url.
          Also, "<managed-server1-ip>:<managed-server1-port>,<managed-server2-ip>:<managed-server2-port>" as your cluster's listen address.
          The following lonks may helpful to you
          http://e-docs.bea.com/wls/docs81/cluster/features.html#1007508
          http://e-docs.bea.com/wls/docs81/cluster/failover.html#1008850
          http://e-docs.bea.com/wls/docs81/cluster/load_balancing.html#1045512
          Regards
          Prasanna Yalam

Domain name/"primary DNS" name/mail host name/mail server name/mx name

Hi,
I have registered my domain name (N.com) with an external dns server and created an MX record (mail.N.com) for it as well. My server sits behind a router and internet traffic is port(80)-forwarded to my server's fixed internal ip address (I can access my webpages from the internet just fine). When I initially setup the server I was asked to give it a "primary DNS name." I naturally assumed that N.com was supposed to be entered here, but that just caused all kinds of problems (though I still do not understand why). So I reinstalled and currently have server.N.com as the primary DNS name of the server (although it shows up in Server Admin and Workgroup Manager as server.local--why is that?)
I have had no luck getting any email from the internet with the default settings in mail services (domain name of N.com and host name of server.N.com). Having no luck with the defaults, I assumed that maybe I needed to change the host name to "mail.N.com" so that it matched the MX record. I also changed the user preferences in Workgroup Manager to receive mail from mail.N.com--but still no luck. Could someone tell me what I am doing wrong and how come none of the Mail Service literature mentions anything about what a mail "host name" is supposed to be? Is it supposed to be the same as the "Mail Server" name that Workgroup Manager asks each user for? and is it the same as the MX record name?
I am just needing some help to connect all these variously named, but undefined, dots.
Also, how come I can access webmail from the internet using www.N.com/webmail but can't do it from any computers within my physical network (I have to use IPaddress/webmail or server.local/webmail)?
Also, should my user email addresses be [email protected] or [email protected] (which is the current default)?
Thanks in advance and good luck!
John

I had been told by a friend that ... regular email coming in from the internet would go through port 80
Unfortunately you were misinformed.
My MX record needs the extra subname (such as "mail") in front of the domain name, right?
Not at all. A MX record for 'N.com' is entirely valid (expected, in fact).
An MX record tells remote mail servers where to send mail for any domain/subdomain. If you think about it, let's say you worked for Apple and you wanted people to be able to send email to [email protected], well then you need an MX record for 'apple.com'.
You can see if you dig MX apple.com that they actually have 9 MX records, but the point still stands.
Now, you might also have MX records for subdomains so that you can have [email protected], [email protected], [email protected], etc. Each of these subdomains would need a separate MX record.
So, in general, for any set of email addresses @[anything.]domain.com you have a MX record telling remote mail servers where to send that mail.
Of course the MX record name is mail.N.com, but I think you are implying that the in the Mail settings of Server Admin where it asks for domain name all I need to put there is the N.com, right?
What I mean is that you need to set the domain name to whatever domain name you expect to receive mail at.
If you want users to have email addresses in the form of [email protected], then you enter 'N.com'.
If you want users to have email addresses in the form of [email protected] then you enter 'mail.N.com'.
If you want both forms, enter one in the main domain and add the other(s) in the Advanced -> Hosting -> Local Host Aliases section.
Is there anything in the Mail settings that needs to have the full MX record name (mail.N.com)?
Yes. Either the 'domain name' or the 'Local Host Aliases' needs to contain the same thing as your MX record. That's because the MX record tells remote servers to send mail to this machine, but the machine won't accept the mail unless it is configured to do so.
What is the "host name" supposed to be?
This should be the name that remote servers see when this machine tries to send outgoing mail.
Ideally this should match the reverse DNS of your IP address, that way when it connects to a remote server it says "Hi, I'm $hostname" and the remote server can lookup the machine's IP address and see the same result. This will reduce the problem of remote servers thinking you're sending them spam.
If you only have a single IP address then this probably needs to be something like 'N.com'. If you don't have control over your reverse DNS then you're going to run into a problem.
Also in the user accounts (in Workgroup Manager) what should the "mail server name" be? Is it the host name or the domain name?
Off hand, I don't know.
Also, why are all the default user email addresses [email protected] (the name I gave the server at setup)? Why doesn't the user accounts create [email protected] as the default address?
Presumably because the mail server says it is 'server.N.com' and therefore any accounts on that machine would be [email protected]. Changing the domain name (as above) should fix that.

EJB lookup works in embedded OC4J but not stand-alone

I use JDeveloper 9.0.3.1 to develop a combined web & EJB application. The EJBs are deployed on a 9.0.3 iAS server on a different machine.
The EJB lookup is in a helper class that's called by the servlet. For EJB lookup, I use the class "com.evermind.server.rmi.RMIInitialContextFactory" as described by the OC4J servlet guide (http://download-uk.oracle.com/docs/cd/A97688_10/generic.903/a97680/develop.htm#1003973). So my code looks like this:
Hashtable map = new Hashtable();
map.put(Context.INITIAL_CONTEXT_FACTORY, "com.evermind.server.rmi.RMIInitialContextFactory");
map.put(Context.PROVIDER_URL, "ormi://server:port/app-business");
map.put(Context.SECURITY_PRINCIPAL, "admin");
map.put(Context.SECURITY_CREDENTIALS, "password");
map.put("dedicated.rmicontext", "true");
InitialContext context = new InitialContext(map);
Object thing = context.lookup("MyEjb");
EJBHome feedback = (EJBHome)PortableRemoteObject.narrow(thing, "com.company.ejb.MyEjbHome");
The last insert into the map (map.put("dedicated.rmicontext", "true");) wasn't mentioned neither in the servlet guide nor the EJB one. But without it, I'd always get some "domain is null" error.
Now this code above works flawlessly in the embedded OC4J of JDeveloper. But it always hangs in the "Object thing = context.lookup("MyEjb");" line when executed either in a stand-alone OC4J (the one from JDeveloper launched through the batch file in [JDeveloper directory]\jdev\bin\start_oc4j.bat) or in the OC4J in iAS.
Does anybody know why this happens or how to fix it? I know that there are other ways of EJB lookup in servlets but since we also have plain Java classes for testing purposes that access EJBs, this seemed like the simplest thing to do for both Java classes and servlets.

I was referring to the java client case where this optoin is not required. In general for web clients one doesn't use the rmi initial context to find the bean. This is because the bean and the webclient are both packaged in the same application. One defines the ejb-ref in the web.xml for the bean and then looks it using the default initial context. Something like
Context ctx = new InitialCOntext();
ctx.lookup("bean ejb-ref-name");
However if you must use rmi for accessing the bean from the web client then the dedicated connection property is required.
For java clients you would setup the rmi context environment.
Since you are trying to use a common utility class for both java and webclient I would recommend that you look into packaging the java client as a j2ee application client. Basically in that case you would be able to define a ejb-ref in the aplication-client.xml and use the default initial context in code for both the cases.
Hope that helps
Dhiraj

EJB lookup returned stub from a different classloader.

I've written an EJB for doing authenitcation. This EJB is accessed by an security-mbean (BEA's login module).
- The EJB is deployed in an EAR.
- The EJB-stubs are extracted and is included as part of the MBEAN Jar.
When I hit a webapp causing the EJB lookup to occur, the stub object returned is created by the webapp's classloader. This cause a ClassCastException when trying to cast the returned home interface into the home interface of the MBEAN's classloader.
Note that all this is happening on the same BEA server running WLS8.1.
I don't want to put the EJB jar on the system classpath so I can redeploy the EJB, the Application's EAR and the MBEAN to a cluster.
-alex

Robert Greig <[email protected]> wrote:
Thanks for responding to my question, which newsgroup is more appropriate for
my line of question?
But before I move this thread, I would like to add:
I've already handled the recursion problem on top of the ejb-lookup before JNDI
becomes avaliable (while doing server startup) problem.
The advantage with the EJB model is this. By changing the host/port configuration,
I can switch between a local-authentication server or a remote provide authentication
server network configuration.
If I were to include the necessary classes in the mbean JAR from our application,
there maybe resources issues since I now have 2 classloaders loading my server-portion
of classes. Not sure how that will workout with resources and all. This model
has the disadvantage of any classes I have in the mbean JAR will require updates
outside my EAR. This wroks against the EAR deployment model.
-alex
Alex Cheung wrote:
I've written an EJB for doing authenitcation. This EJB is accessedby an security-mbean (BEA's login module).
- The EJB is deployed in an EAR.
- The EJB-stubs are extracted and is included as part of the MBEANJar.
This isn't a good approach. You are pretty much stuffed mainly for the
reasons you outline.
Also note that if you continue to go down this road you will have to
handle the potential recursion (i.e accessing an EJB will invoke a
security call to your provider!).
Why do you need to implement this as an EJB? The main advantages of EJBs
are security and container managed transactions neither of which is
relevant here surely?
Robert

Ejb lookup and oc4j container startup

I have a servlet that is set to auto start on container startup that does a lookup on a ejb in another container. jndi locates the ejbhome but fails to load the class for the ejb home. It knows the name of the class but the classloader cannot find it. This same servlet works fine after the container has started up.
It feel the problem is a classloader problem or a restriction by oracle. Any help is appreciated.
Why can't the OC4J default application that auto starts servlets at container startup not find the ejbhome class?
David Jacobs

We have previously done it, but it only worked when we are using in 10g AS R1. But now, we are porting to R2, but it doesn't work. We didn't do any other things, just a simple lookup using the url, context factory. In R1 it worked fine, in R2, we are encountering the same, NameNotFoundException. Hope someone could help us also with this. Another thing, when we are using ormi://<rest of url>, it doesn't return any exception and our application hangs, but if we use opmn:ormi://<rest of URL>, that's the time it throws the NameNotFoundException. In R1, we didn't have any problems.

Cluster network name resource error 1196

I have a two-node Hyper-V 2012 Core Server cluster. Initially, the cluster was working fine and I was able to do live migrations with no problem. I first realized that something had gone wrong when I was attempting to move a VM's storage to the Cluster Disk.
When I right click on the VM and select Move > Virtual Machine Storage, the bottom left area of the "move storage" screen says "Loading..." forever and never displays the cluster storage. This happened once before and I was able to correct
the issue by taking the Server Name resource offline and doing More Actions > Repair. However, if I try a repair now I receive the following error messages:
"There was an error repairing the active directory object for 'Cluster Name'. Error Code: 0x800713b8. The cluster request is not valid for this object".
I have verified that the cluster name Active Directory object (CCSDCluster) still exists and each node as well as itself has full permissions to the object. What I am not certain of is whether the object is still in the same Organizational Unit as it was
originally. After the failed repair attempt, I am able to bring the Server Name resource back online successfully.
In addition, I am continually receiving cluster event errors indicating the following:
"Cluster network name resource 'Cluster Name' failed registration of one or more associated DNS name(s) for the following reason: The handle is invalid. Ensure that the network adapters associated with dependent IP address resources are configured
with at least one accessible DNS server."
Each of the two nodes has multiple NICs configured. One for management, one for guest OSs (Hyper-V virtual switch is configured to use this NIC and not allow management OS access), and two for iSCSI (for redundancy and multi-path). Live migration is set
to only use the management NICs. The management NIC on both nodes is configured with valid DNS servers. The CCSDCluster DNS entry has full permissions for both nodes and itself.
Any suggestions on what to try would be greatly appreciated. Thank you!

Could my issues be related to a problematic network configuration? Please see the ipconfig below from one of my Hyper-V host servers.
Ethernet adapter Ethernet 8 = iSCSI
Ethernet adapter Ethernet 6 = iSCSI
Ethernet adapter Ethernet 2 = Management interface
10.1.3.19 = Management IP address, 10.1.3.3 = IP address of Cluster Name Object
Tunnel adapter Local Area Connection *11 = ?? Not sure what this one is
Windows IP Configuration
Host Name . . . . . . . . . . . . : CCSDVM01
Primary Dns Suffix . . . . . . . : clairton.local
Node Type . . . . . . . . . . . . : Hybrid
IP Routing Enabled. . . . . . . . : No
WINS Proxy Enabled. . . . . . . . : No
DNS Suffix Search List. . . . . . : clairton.local
Ethernet adapter Ethernet 9:
Media State . . . . . . . . . . . : Media disconnected
Connection-specific DNS Suffix . :
Description . . . . . . . . . . . : HP NC552SFP Dual Port 10GbE Server Adapter #4
Physical Address. . . . . . . . . : 28-92-4A-AF-46-34
DHCP Enabled. . . . . . . . . . . : Yes
Autoconfiguration Enabled . . . . : Yes
Ethernet adapter Ethernet 8:
Connection-specific DNS Suffix . :
Description . . . . . . . . . . . : HP NC552SFP Dual Port 10GbE Server Adapter #3
Physical Address. . . . . . . . . : 28-92-4A-AF-61-68
DHCP Enabled. . . . . . . . . . . : No
Autoconfiguration Enabled . . . . : Yes
IPv4 Address. . . . . . . . . . . : 172.16.1.12(Preferred)
Subnet Mask . . . . . . . . . . . : 255.255.255.0
Default Gateway . . . . . . . . . : 172.16.1.1
NetBIOS over Tcpip. . . . . . . . : Enabled
Ethernet adapter Ethernet 7:
Media State . . . . . . . . . . . : Media disconnected
Connection-specific DNS Suffix . :
Description . . . . . . . . . . . : HP NC552SFP Dual Port 10GbE Server Adapter #2
Physical Address. . . . . . . . . : 28-92-4A-AF-61-6C
DHCP Enabled. . . . . . . . . . . : Yes
Autoconfiguration Enabled . . . . : Yes
Ethernet adapter Ethernet 6:
Connection-specific DNS Suffix . :
Description . . . . . . . . . . . : HP NC552SFP Dual Port 10GbE Server Adapter
Physical Address. . . . . . . . . : 28-92-4A-AF-46-30
DHCP Enabled. . . . . . . . . . . : No
Autoconfiguration Enabled . . . . : Yes
IPv4 Address. . . . . . . . . . . : 172.16.1.11(Preferred)
Subnet Mask . . . . . . . . . . . : 255.255.255.0
Default Gateway . . . . . . . . . : 172.16.1.1
NetBIOS over Tcpip. . . . . . . . : Enabled
Ethernet adapter Ethernet 3:
Media State . . . . . . . . . . . : Media disconnected
Connection-specific DNS Suffix . :
Description . . . . . . . . . . . : Broadcom BCM5709C NetXtreme II GigE (NDIS VBD Client) #55
Physical Address. . . . . . . . . : E4-11-5B-B9-3A-9C
DHCP Enabled. . . . . . . . . . . : Yes
Autoconfiguration Enabled . . . . : Yes
Ethernet adapter Ethernet 2:
Connection-specific DNS Suffix . :
Description . . . . . . . . . . . : Broadcom BCM5709C NetXtreme II GigE (NDIS VBD Client) #58
Physical Address. . . . . . . . . : E4-11-5B-B9-3A-A2
DHCP Enabled. . . . . . . . . . . : No
Autoconfiguration Enabled . . . . : Yes
IPv4 Address. . . . . . . . . . . : 10.1.3.19(Preferred)
Subnet Mask . . . . . . . . . . . : 255.255.255.0
IPv4 Address. . . . . . . . . . . : 10.1.3.3(Preferred)
Subnet Mask . . . . . . . . . . . : 255.255.255.0
Default Gateway . . . . . . . . . : 10.1.3.100
DNS Servers . . . . . . . . . . . : 10.1.3.10
10.1.3.39
NetBIOS over Tcpip. . . . . . . . : Enabled
Tunnel adapter Local Area Connection* 11:
Connection-specific DNS Suffix . :
Description . . . . . . . . . . . : Microsoft Failover Cluster Virtual Adapter
Physical Address. . . . . . . . . : 02-AD-4B-6D-A9-FB
DHCP Enabled. . . . . . . . . . . : No
Autoconfiguration Enabled . . . . : Yes
IPv4 Address. . . . . . . . . . . : 169.254.2.92(Preferred)
Subnet Mask . . . . . . . . . . . : 255.255.0.0
Default Gateway . . . . . . . . . :
NetBIOS over Tcpip. . . . . . . . : Enabled

Prefer DNS name to last known IP

One of the most annoying behaviors of ARD is that is seems to care only about the IP of the client. I frequently lose connections to clients in our DHCP environment when the IP changes. ARD already knows the DNS name of the client, but it does not seem to even try to do a DNS lookup to try to reconnect when the IP changes. Does anyone know of a hack or other method to have ARD do a DNS lookup when it cannot contact a client?
Thanks!

Get Info on the client. Edit the info. Delete the IP address. You may, or may not, need to make an edit of some sort in the DNS field (so it registers a change). Save the new info.
This should force ARD to prefer the DNS name over the IP address.

9.0.3 EJB Lookup

Using the latest 9.0.3 Application Server,
whenever I try to lookup an EJB (using RMI
and its JNDI name), my application completely
hangs at the lookup.
Code Example (Works on every other Application Server):
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
String JNDIName = "ess.sis.server.Dispatcher";
DispatcherHome homeInterface = null;
SIDispatcherV4 remote = null;
Properties h = new Properties();
h.put(Context.INITIAL_CONTEXT_FACTORY, factory );
h.put( Context.PROVIDER_URL, sURL );
h.put( Context.SECURITY_PRINCIPAL, sUSERNAME );
h.put( Context.SECURITY_CREDENTIALS, sPASSWORD );
InitialContext ctx = new InitialContext(h);
try
//LINE BELOW HANGS INDEFINATLY.
homeInterface = (DispatcherHome) ctx.lookup( JNDIName );
remote = (SIDispatcherV4) homeInterface.create( );
catch (Exception e)
e.printStackTrace();
Is there a setting that must be activated to
allow this action on the new application server ?
Cheers,
Ronan

Ronan (and Jeff :-),
Let me try again (since Ronan ignored my previous reply -- and still hasn't supplied the details I asked for [believe me, if you give me the information, I'm quite sure I can help you!]) because I'm in agony watching you guys going round in circles!
Ronan, the following applies to a stand-alone java client (a normal java class with a "main()" method that you execute from a command prompt using the "java" command) -- not deployed to OC4J. Client code should be (correcting your original code):
// lookup name is content of <ejb-name> from "ejb-jar.xml"
String JNDIName = "ess.sis.server.Dispatcher";
Properties h = new Properties();
h.put(Context.INITIAL_CONTEXT_FACTORY,
"RMIInitialContextFactory");
// 'your_app_name' is value of "name" attribute from
// relevant "application" element in "server.xml" file.
h.put( Context.PROVIDER_URL,
"ormi://<your_host>/<your_app_name>");
// use the "admin" user if you don't want to spend time
// configuring the "principals.xml" file
h.put( Context.SECURITY_PRINCIPAL, "admin");
// use the password you defined when installing OC4J
h.put( Context.SECURITY_CREDENTIALS,"<password>");
InitialContext ctx = new InitialContext(h);
Object homeObj = ctx.lookup(JNDIName)Long postings are being truncated to ~1 kB at this time.

EJB lookup in a cluster (DNS names to servers mapping)

Similar Messages

Maybe you are looking for