JNDI clustering

When I do a:
          Context context = new InitialContext();
          In a clustered environment, will that connect to the cluster JNDI or the
          localhost JNDI?
          We need to make sure whenever our stateless session beans get a handle
          to the context, it is the localhost JNDI, Is there a way to do this if
          the answer to the above question is the cluster?
          The problem we have is we are binding local objects into JNDI, each
          server in the cluster needs to get references to one of these objects
          (the stateless session bean needs to do a lookup and get a reference).
          So our code looks something like this:
          public class MyStatelessSessionBean implements SessionBean
          public void myRemoteCall(...) throws RemoteException
          Context context = new InitialContext();
          MyLocalObject object = (MyLocalObject)context.lookup("MyObject");
          We need to ensure that the MyLocalObject that is returned is one from
          the localhost, not one in the cluster.
          thanks,
          -chris
          

On the server if you do a Context ctx = new InitialContext() it will return you
          local initial context. So, your lookup will be local.
          Instead if you use a dns cluster name then the first server name that dns server
          returns will be used to connect and get an initial context if you are on a
          standalone client.
          Environment env = new Environment();
          env.setProviderUrl("t3://fooclust:7001");
          Context ctx = env.getInitialContext()
          When you do a lookup in the jndi tree, what you get back depends on the server
          configuration and doesn't depend on where you look if you are in a homogenous
          cluster.
          - Prasad
          Chris Humphrey wrote:
          > Thanks for the reply. We are currently using the Referenceable interface to
          > store and get objects from jndi, so the objects are not serializable.
          >
          > It sounds like I have a problem to fix here, let me recap to make sure I
          > understand. If the bean was accessed with a DNS cluster for the context, I
          > will get back a reference to the cluster, if the bean was accessed using
          > 'localhost' I will get back a reference to the jndi running locally.
          >
          > This was what I was afraid of. Is there a way to forceably get an initial
          > context to the localhost jndi?
          >
          > Thanks again for the reply.
          > -chris.
          >
          > Wei Guan wrote:
          >
          > > Comments inline
          > >
          > > --
          > > Hope it helps.
          > >
          > > Cheers - Wei
          > >
          > > "Chris Humphrey" <[email protected]> wrote in message
          > > news:[email protected]...
          > > > When I do a:
          > > >
          > > > Context context = new InitialContext();
          > > >
          > > > In a clustered environment, will that connect to the cluster JNDI or the
          > > > localhost JNDI?
          > > In cluster, JNDI tree is "replicated JNDI tree". Every server in the
          > > cluseter has the same "replicated JNDI tree".
          > > " Context context = new InitialContext();" will give you reference to the
          > > JNDI tree in the same JVM as the caller's.
          > >
          > > >
          > > > We need to make sure whenever our stateless session beans get a handle
          > > > to the context, it is the localhost JNDI, Is there a way to do this if
          > > > the answer to the above question is the cluster?
          > > >
          > > > The problem we have is we are binding local objects into JNDI, each
          > > > server in the cluster needs to get references to one of these objects
          > > > (the stateless session bean needs to do a lookup and get a reference).
          > > >
          > > > So our code looks something like this:
          > > >
          > > > public class MyStatelessSessionBean implements SessionBean
          > > > {
          > > > public void myRemoteCall(...) throws RemoteException
          > > > {
          > > > Context context = new InitialContext();
          > > > MyLocalObject object = (MyLocalObject)context.lookup("MyObject");
          > > > }
          > > > }
          > > >
          > > > We need to ensure that the MyLocalObject that is returned is one from
          > > > the localhost, not one in the cluster.
          > > If your "MyObject" is a serializable object, serializable objects binding
          > > the JNDI tree will be replicated. That is, "MyObject" will have a copy on
          > > each server's JNDI tree. Since you get JNDI reference from localhost, your
          > > lookup("MyObject") will get the copy in the same JVM.
          > >
          > > >
          > > > thanks,
          > > > -chris
          > > >
          

Similar Messages

  • JNDI /Clusters (Rob can u please take a look at this ?)

    Hi ..
    I am trying to bind a custom object (which offcourse implements
    Serializable) in to the JNDI tree..
    The important thing to note here is that the 2 WLS servers are in a
    cluster and I am getting the Initial Context with the with the following
    (default) properties..
    h.put(WLContext.REPLICATE_BINDINGS, "true");
    h.put(WLContext.PIN_TO_PRIMARY_SERVER,"false");
    h.put(Context.PROVIDER_URL, t3://Server1,Server2:7010 );
    Now my custom object's values keep changing Dynamically and hence I am
    doing a unbind and binding as
    Lets say I do this on "Server2"
    MyObject zz = new MyObject( );
    zz.changeFustration_Level("Happy");
    ctx.bind("MyObject",zz);
    //As the program logic changes.... (after some interval....)
    ctx.unbind("MyObject");
    //Update MyObject .....
    MyObject oo = new MyObject( )
    oo.changeFustration_Level("Fustrated");
    ctx.bind("MyObject",oo);
    And I assumed (happily...) that the new Object would get replicated in
    the cluster ....
    (Also resolvedObject: weblogic.jndi.toolkit.ReplicatedWLContext is
    called one of the servers)
    But to my utter disgust the Object (in this case MyObject) is a stale
    copy on the other server.....("Server1")
    And to the best of my knowledge my code is doing everything as described
    in the Docs...
    Could anyone throw any insight in to this ??
    Awaiting reply from all the JNDI gurus...
    Thanxs,
    Naggi

    That thread has been noticed before. It all started with a bad suggestion that no one noticed and corrected. Then the post got slightly scrambled in the new forums. Finally, the original poster added some special root powers and a mess ensued.

  • JNDI lookup using a proxy server

    Hi,
    I'm trying to use a weblogic server as a proxy server for a cluster of
    weblogic servers. The static and dynamic content delivery works fine but
    when I try to do a JNDI lookup from an applet client, I get a name not
    resolved exception. I guess, it happens because the proxy server tries to
    resolve the name from it's JNDI tree, whereas I expect it to forward this
    request to one of the servers in the cluster. Is this possible? If it is, do
    I need to set up some parameters in the proxy server? If not, can I do this
    myself by writing a name resolution class which in turn will forward the
    request to one of the servers in the cluster?
    Any help would be greatly appreciated.
    thanks
    - sanjay

    when u say "cluster of weblogic servers"..u mean the weblogic`s inbuilt
    cluster..
    the reason i ask is weblogic`s cluster do provide JNDI clustering too by using
    cluster-wide naming tree..
    Sanjay Aggarwal wrote:
    Hi,
    I'm trying to use a weblogic server as a proxy server for a cluster of
    weblogic servers. The static and dynamic content delivery works fine but
    when I try to do a JNDI lookup from an applet client, I get a name not
    resolved exception. I guess, it happens because the proxy server tries to
    resolve the name from it's JNDI tree, whereas I expect it to forward this
    request to one of the servers in the cluster. Is this possible? If it is, do
    I need to set up some parameters in the proxy server? If not, can I do this
    myself by writing a name resolution class which in turn will forward the
    request to one of the servers in the cluster?
    Any help would be greatly appreciated.
    thanks
    - sanjay

  • Non-replicated resource

    I am fighting weblogic's RMI/JNDI/clustering mechanism. How do I get
              'normal' non-failover, globally available RMI objects?
              I have an external device to which exactly one of the servers in a
              cluster is connected by a serial cable.
              If that server goes down, the users physically move the serial cable,
              and that server should now talk to the device.
              The device is output only. The information to be sent to the device is
              stored in the database.
              I implemented the component that talks to the device -- the 'Talker' --
              as an RMI object which rebinds itself only when it detects the device on
              its serial port. Note the use of rebind rather than bind. This should
              replace any binding from a peer Talker in the cluster that was previous
              communicating with the device.
              It is not working at all.
              Before rebinding, a Talker first looks any bound peer. It never finds
              one, even if a peer has bound itself in the registery.
              Then, when it rebinds, I get cryptic log messages:
              <ConflictHandler> ConflictStart deviceTalker:com.xxx.dom.impl.TalkerImpl
              (from [email protected]:[80,80,7002,7002,-1])
              Details:
              I did NOT rmic with -clusterable.
              I did extend weblogic.rmi.server.UnicastRemoteObject.
              

    I've tried. The documentation is not very clear on how to do this.
              What I did is:
              1. rmic without the -clustered option
              2. I create the context and as follows:
              Hashtable env = new Hashtable();
              env.put(weblogic.jndi.WLContext.PIN_TO_PRIMARY_SERVER, "true");
              i_ctx = new InitialContext(env);
              i_ctx.rebind(JNDI_NAME, this);
              Note that the Talker component doesn't bind() on startup. Instead, it
              rebinds() whenever it determines that it is connected to the device.
              Alexander Petrushko wrote:
              > Mark Mendel wrote:
              >
              > > I am fighting weblogic's RMI/JNDI/clustering mechanism. How do I get
              > > 'normal' non-failover, globally available RMI objects?
              >
              > Have you followed the procedure for looking up pinned objects? See
              > http://www.weblogic.com/docs45/classdocs/API_cluster.html#explaintargeted
              >
              > Cheers
              >
              > Alex
              >
              > --
              > mailto:[email protected] // Consulting services available
              

  • Clustered Remote Object Bound To JNDI

    Hello,
              I'd like to know if Weblogic 5.1 (service pack 10) supports
              clusterable remote objects.
              What I am trying to do is to implement a remote object that extends
              UnicastRemoteObject interface and bind it to JNDI. When the client
              needs the services, it will lookup that object in JNDI and call its
              methods remotely.
              What happens to my test implementation is that it all works well when
              I run Weblogic locally, but if it is run on the remote server, it
              gives me the exception to the extend that the stub for my remote
              object is not found (although it is in the classpath). I understand
              that this is because I should be using weblogic.rmic instead of
              regular rmic, is that right?
              Thanks,
              Eugene Kononov.
              

    Hello,
              I'd like to know if Weblogic 5.1 (service pack 10) supports
              clusterable remote objects.
              What I am trying to do is to implement a remote object that extends
              UnicastRemoteObject interface and bind it to JNDI. When the client
              needs the services, it will lookup that object in JNDI and call its
              methods remotely.
              What happens to my test implementation is that it all works well when
              I run Weblogic locally, but if it is run on the remote server, it
              gives me the exception to the extend that the stub for my remote
              object is not found (although it is in the classpath). I understand
              that this is because I should be using weblogic.rmic instead of
              regular rmic, is that right?
              Thanks,
              Eugene Kononov.
              

  • JNDI lookup in Multitired clustered architecture

    Hi alll,
    We are in the design phase of a multi tired clustered application.
    This application has jsp/servelet cluster and business application cluster. We are confused abt the JNDI lookup logic need to use in servlet cluster.
    Can anyone help me to find out an optimal solution for jndi which gives the complete advantage of load balancing and esp. fail over
    of clustering
    Thanks in advance
    gokul

    Hi,
              First of all, I would like to clear the point that load balancing can be done within a cluster not among the cluster.
              So if you have more than one cluster the load balancing
              would be at each cluster level.
              For JNDI lookup you have to provide cluster address i.e combination of Managed Server Address e.g:
              t3://localhost:7001,localhost:8001,localhost:9001...etc
              Thanks,
              Qumar

  • JNDI lookup in a clustered environment

              Hi,
              I have deployed my application on a clustered environment containing one admin
              server and two managed servers. The application is deployed on the managed servers.
              We have a DNS for these two managed servers.
              Currently, I am using t3:localhost:portnumber to get an initial context for JNDI
              lookup.
              But this will not provide perfect load balancing, since each server side component
              (EJB) will lookup for a needed component on the same server itself.
              Is there any way to get the Initial Context by using the DNS name?
              I tried t3:DNSName (Our DNS is created till the port number. e.g. http://DNS points
              to http://Managedserver1:portnumber and http://Managedserver2:portnumber
              It works, but is random. Sometimes, it gives me the following exception:
              Root exception is java.net.ConnectException: t3://DNS Name: Destination not
              reachable using: 't3'
              Could someone help me figure out a way to do this in a better way?
              Thanks,
              Aparna
              

              If you don't want or have money to use external load balancers you should be using
              the DNS entry as t3://server1:port,server2:port (you indicated you are using http.
              Also make sure your network elements between the client and the servers allow
              t3 protocol if there is packet filtering).
              S
              "Aparna" <[email protected]> wrote:
              >
              >Hi,
              >
              >I have deployed my application on a clustered environment containing
              >one admin
              >server and two managed servers. The application is deployed on the managed
              >servers.
              >We have a DNS for these two managed servers.
              >
              >Currently, I am using t3:localhost:portnumber to get an initial context
              >for JNDI
              >lookup.
              >But this will not provide perfect load balancing, since each server side
              >component
              >(EJB) will lookup for a needed component on the same server itself.
              >
              >Is there any way to get the Initial Context by using the DNS name?
              >
              >I tried t3:DNSName (Our DNS is created till the port number. e.g. http://DNS
              >points
              >to http://Managedserver1:portnumber and http://Managedserver2:portnumber
              >
              >It works, but is random. Sometimes, it gives me the following exception:
              > Root exception is java.net.ConnectException: t3://DNS Name: Destination
              > not
              >reachable using: 't3'
              >
              >Could someone help me figure out a way to do this in a better way?
              >
              >Thanks,
              >Aparna
              

  • Can a single JNDI tree be shared by multiple, non-clustered servers?

    We have a situation involving multiple, but non-clustered, WebLogic servers,
    say Server 1 and Server 2. A client needs to access an EJB which happens to
    be located only on Server 2. We would prefer that the client not need to
    know the exact location of the EJB, but use the naming service hosted by a
    designated server (Server 1) to locate the EJB.
    At a minimum, this would require that both servers bind their respective
    EJB's into a common, shared JNDI naming tree hosted by Server 1.
    Is this possible?
    John Armstrong
    WebLink Wireless, Inc.
    [email protected]

    hi,
    I have not tried this, althoug with the webogic rmi and weblogic jndi
    implementation I would guess this is straightforward.
    Binding remote objects in WL JNDI binds the stubs only, not the object
    itself. The worst case scenario here would be for you to implement proxies
    and bind them in JNDI, although I would guess that binding the home object
    should give you the required functionality.
    Just try it, rebind the home object on a different server than the one the
    bean is deployed on and test. My guess is that it will work.
    (I repeat : I have not tested this my self, this is just a suggestion)
    Regards,
    Anders M.
    John N. Armstrong <[email protected]> skrev i
    meldingsnyheter:3a93faba$[email protected]..
    We have a situation involving multiple, but non-clustered, WebLogicservers,
    say Server 1 and Server 2. A client needs to access an EJB which happensto
    be located only on Server 2. We would prefer that the client not need to
    know the exact location of the EJB, but use the naming service hosted by a
    designated server (Server 1) to locate the EJB.
    At a minimum, this would require that both servers bind their respective
    EJB's into a common, shared JNDI naming tree hosted by Server 1.
    Is this possible?
    John Armstrong
    WebLink Wireless, Inc.
    [email protected]

  • Clustering JNDI

    Hi,
    As I'm using clustering of weblogic 7.0, and I know I cannot use ServletContext
    for placing objects as it's not replicated via all the servers, I wondered if
    I could use Context for that.
    If I place objects in the JNDI Context using bind, and I update them using rebind,
    will those object and changes be clustered to every node? What should I do for
    changes to be propagated?
    Does the server's perfomance suffer too much from changing use of JNDI instead
    of ServletContext?
    Is there any problem for concurrent access to objects in JNDI Context?
    I know there's a product called Coherence by Tangosol for using ServletContext
    in clustered environments, but I cannot use it as my client does not wanna hear
    about more products to be bought. :.-(
    Thanks

    "Ignacio Sanchez" <[email protected]> wrote in message news:<[email protected]>...
    Thanks for your reply,
    Anyway, I see one big problem, which is included in that page, in the warning.
    Here it is: "Do not use the cluster-wide JNDI tree as a persistence or caching
    mechanism for application data. Although WebLogic Server replicates a clustered
    server instance's JNDI entries to other server instances in the cluster, those
    entries are removed from the cluster if the original instance fails".
    So, this does not seem to be the solution.
    "SIPTech Community" <[email protected]> wrote:
    "Ignacio Sanchez" <[email protected]> wrote:
    Hi,
    As I'm using clustering of weblogic 7.0, and I know I cannot use ServletContext
    for placing objects as it's not replicated via all the servers, I wondered
    if
    I could use Context for that.
    If I place objects in the JNDI Context using bind, and I update them
    using rebind,
    will those object and changes be clustered to every node? What should
    I do for
    changes to be propagated?
    Does the server's perfomance suffer too much from changing use of JNDI
    instead
    of ServletContext?
    Is there any problem for concurrent access to objects in JNDI Context?
    I know there's a product called Coherence by Tangosol for using ServletContext
    in clustered environments, but I cannot use it as my client does not
    wanna hear
    about more products to be bought. :.-(
    ThanksHi,
    Every server instance in a cluster maintains Clusterwide JNDI Tree
    (for clustered
    objects).
    Binded objects in any cluster server instance will be replicated
    in other
    server instances just after successful
    binding.
    Pls. view this URL for more details :
    http://edocs.bea.com/wls/docs70/cluster/features.html#1007508
    Prema
    Many Application Servers make use of caching for the JNDI Tree when
    you have more than 1 clone (vertical or horizontal).
    If you make use of JNDI in order to store "constant" information, you
    need to use just bind /rebind and lookup... but if you want update
    frequently a JNDI object you will have to clean this cache by setting
    the appropriate property in the JNDI Environment... because you could
    experience a dirty (old) value by making a simple lookup.
    A sample using IBM WebSphere:
    public static final Object lookupEnvEntry(InitialContext ic, String
    entryName, boolean clearCache) {
         try {
              if (clearCache) {
                   Hashtable env = ic.getEnvironment();
                   env.put("com.ibm.websphere.naming.jndicache.cacheobject",
    "cleared");
                   ic = new InitialContext(env);
              return (String)ic.lookup(entryName.trim());
         } catch (Exception ex) {
              System.out.println("Initial Context: Error retrieving Environment
    Entry: " + entryName.trim() + " - " + ex.getMessage());
              return null;

  • Clustering objects in jndi

    Hi all,
    I am building a solution using weblogic version 6.1sp1.
    I am needing to share an object around a set of clusters to tell each cluster
    about the services in the other clusters.
    The first solution I thought of was to create a javax.naming.Reference to bind
    into each of the clusters' JNDI tree and have it replicated throughout the tree.
    After reading about the behavior of these types of objects, I am not sure how
    to proceed.
    It appears from the docs that if the server that bound the object into jndi goes
    down the object will be unbound from all of the other servers in the cluster.
    Please inform if my interpretation of the docs was incorrect.
    The solution I am working for now is to not replicate the objects, and to create
    a startup class for all of the servers, when they come up they will create and
    bind their own versions.
    Also, if anyone has another solution to this problem please reply ...

    Your solution sounds right. Don't forget to set REPLICATE_BINDINGS to 'false' :
    http://e-docs.bea.com/wls/docs61/javadocs/weblogic/jndi/WLContext.html
    Chris Humphrey <[email protected]> wrote:
    Hi all,
    I am building a solution using weblogic version 6.1sp1.
    I am needing to share an object around a set of clusters to tell each cluster
    about the services in the other clusters.
    The first solution I thought of was to create a javax.naming.Reference to bind
    into each of the clusters' JNDI tree and have it replicated throughout the tree.
    After reading about the behavior of these types of objects, I am not sure how
    to proceed.
    It appears from the docs that if the server that bound the object into jndi goes
    down the object will be unbound from all of the other servers in the cluster.
    Please inform if my interpretation of the docs was incorrect.
    The solution I am working for now is to not replicate the objects, and to create
    a startup class for all of the servers, when they come up they will create and
    bind their own versions.
    Also, if anyone has another solution to this problem please reply ...--
    Dimitri

  • How many clusterable things in JNDI tree?

    We're starting to try out clustering in WL 5.1 (sp8 on WinNT for now) We're
              seeing some odd behavior and we're wondering if the cause is related to JNDI
              replication messages. If we deploy just a couple of clusterable rmi
              services on two servers in a cluster, things seem to work ok. They maintain
              a connection to each other and can be viewed with the Admin console.
              However, if we deploy about 200 clusterable things (mostly EBs that by
              default have clusterable homes) then things get weird. If we just bring up
              one server, things aren't too bad. But when we bring up the second server,
              communication between the two servers is only established briefly before we
              get a "Timed out server" or something worse.
              Someone thought they had read a news post saying that it's not good to have
              large number of clusterable things in the JNDI tree because of all the
              replication messages caused by this. True? How many is too many? Thanks.
              

    JNDI replication messages only occur when the state of the JNDI tree changes --
              which is normally at server startup or shutdown. I would recommend pursuing
              this with support as I have been at sites that deployed more EJBs in a cluster
              than this without those problems.
              Joe Herbers wrote:
              > We're starting to try out clustering in WL 5.1 (sp8 on WinNT for now) We're
              > seeing some odd behavior and we're wondering if the cause is related to JNDI
              > replication messages. If we deploy just a couple of clusterable rmi
              > services on two servers in a cluster, things seem to work ok. They maintain
              > a connection to each other and can be viewed with the Admin console.
              >
              > However, if we deploy about 200 clusterable things (mostly EBs that by
              > default have clusterable homes) then things get weird. If we just bring up
              > one server, things aren't too bad. But when we bring up the second server,
              > communication between the two servers is only established briefly before we
              > get a "Timed out server" or something worse.
              >
              > Someone thought they had read a news post saying that it's not good to have
              > large number of clusterable things in the JNDI tree because of all the
              > replication messages caused by this. True? How many is too many? Thanks.
              

  • Re: Clustering & JNDI

    The JNDI documentation says that only Remote objects can be replicated.
              Excerpts from WL jndi doc:
              Note that you cannot reliably use the naming service to replicate objects.
              For example, if you bind an RMI object "myReplicatedObject" into a
              cluster-wide JNDI tree, and "myReplicatedObject" from Server One places a
              serializable -- but not Remote -- object "myDataObject" into the JNDI tree,
              that unRemoted object does not get replicated to other servers in the
              cluster. If Server One fails and a client calls "myReplicatedObject," the
              RMI service will be available, but its "myDataObject" will not.
              If you need to guarantee access to some data that is produced or used by a
              replicated object, you should store the data in a database so that it can be
              reached by any server in the cluster that is hosting the replicated service.
              An example of this is an entity EJBean.
              Sam Jacob
              GiriBabu <[email protected]> wrote in message
              news:[email protected]...
              > Hi,
              >
              > Has anyone tried putting data objects in a JNDI tree, and then setting
              > up the servers in a clustered environment. The behavior I am seeing is
              > strange.
              >
              > Goal: To cache the same data in all the servers in a cluster
              >
              > Proposed Solution: Put the data in the JNDI tree, so that clients can
              > do a lookup and get the data. Since WL does not synchronize data across
              > the servers in a cluster, manually synchronize all the JNDI trees.
              >
              > In the prototype, I have a startup class, which populates the JNDI tree
              > on the URL passed in as an argument. The startup class is registered in
              > the Cluster-wide properties file.
              >
              >
              > After implementing the prototype, this is the bahaviour I see
              >
              > If I kill one of the servers, the client gets an exception, when he does
              > a lookup on a live server.
              >
              > It appears that the dying server is taking away all the JNDI entries
              > that it put in its JNDI tree, from all the other JNDI trees in the
              > cluster too.
              >
              >
              > Am I assuming something wrong? I want to create separate copies of the
              > data objects in each JNDI tree.
              >
              > Thanks ,
              > Vijay
              >
              >
              >
              >
              >
              >
              >
              

    Sam wrote:
              > The JNDI documentation says that only Remote objects can be replicated.
              > Excerpts from WL jndi doc:
              > Note that you cannot reliably use the naming service to replicate objects.
              > For example, if you bind an RMI object "myReplicatedObject" into a
              > cluster-wide JNDI tree, and "myReplicatedObject" from Server One places a
              > serializable -- but not Remote -- object "myDataObject" into the JNDI tree,
              > that unRemoted object does not get replicated to other servers in the
              > cluster. If Server One fails and a client calls "myReplicatedObject," the
              > RMI service will be available, but its "myDataObject" will not.
              >
              > If you need to guarantee access to some data that is produced or used by a
              > replicated object, you should store the data in a database so that it can be
              > reached by any server in the cluster that is hosting the replicated service.
              > An example of this is an entity EJBean.
              >
              >
              > Sam Jacob
              >
              An entity bean is always a good option, but what happens if you want to
              replicate the information in memory for the highest performance?
              

  • JNDI and clustering

    All,
              What happens with naming services in a clustered environment?
              Is the naming service dupilcated across all servers in the cluster or
              are they other approaches?
              How is this configured?
              What is the effect on the client when requesting the initial context ina
              clusteerd environment?
              Thanks in advance
              Aaron
              

    Aaron Robinson wrote:
              > All,
              >
              > What happens with naming services in a clustered environment?
              > Is the naming service dupilcated across all servers in the cluster or
              > are they other approaches?
              >
              Anything you bind into the jndi tree will be replicated.
              > How is this configured?
              Nothing to configure. It is automatic when you start servers in a
              cluster.
              > What is the effect on the client when requesting the initial context ina
              > clusteerd environment?
              None.
              >
              > Thanks in advance
              > Aaron
              Cheers
              - Prasad
              

  • Clustered jndi replication

    Does the replication in clustered jndi also support removing replicate
    bindings
    when you call unbind()?
    When I programmatically bind a value (a String), it is replicated throughout
    the cluster but when I try to unbind it only gets removed on one server.
    Calling rebind with a new value doesn't work either; a message is printed to
    the console saying there is a conflict. Has anyone seen this before and/or
    is there
    something I'm missing.
    This is on weblogic5.1sp4.

    This sounds like a bug.
    I suggest that you file a bug report with our support organization. Be sure
    to include a complete test case. They will also need information from
    you -- please review our external support procedures:
    http://www.beasys.com/support/index.html
    Thanks,
    Michael
    Michael Girdley
    BEA Systems Inc
    "Jonathon Lee" <[email protected]> wrote in message
    news:398b4219$[email protected]..
    Does the replication in clustered jndi also support removing replicate
    bindings
    when you call unbind()?
    When I programmatically bind a value (a String), it is replicatedthroughout
    the cluster but when I try to unbind it only gets removed on one server.
    Calling rebind with a new value doesn't work either; a message is printedto
    the console saying there is a conflict. Has anyone seen this beforeand/or
    is there
    something I'm missing.
    This is on weblogic5.1sp4.

  • JNDI spanning multiple clusters

              Hi!
              I would like to know if it is possible to have multiple
              clusters of servers (potentially in different network
              zones), and still have a global JNDI tree, as if it
              was only one big cluster? What are the considerations
              to meet such a requirement?
              Thanks,
              Guillaume Bedard
              

    Generally speaking, you usually end up with multiple clusters with scenarios
              like this, each separated by a firewall for example.
              Peace,
              Cameron Purdy
              Tangosol, Inc.
              Clustering Weblogic? You're either using Coherence, or you should be!
              Download a Tangosol Coherence eval today at http://www.tangosol.com/
              "Guillaume Bedard" <[email protected]> wrote in message
              news:[email protected]...
              >
              > The client here "needs" to split the whole cluster into
              > sub-clusters in different security zones.
              >
              > Reading through the documentation, it says that it should
              > work as long as all the servers can subscribe and listen
              > to a single multicast IP address, but I am suspecting
              > that the particular network setup (firewalls) might
              > prevent us from having it working...
              >
              > Guillaume
              >
              > "Cameron Purdy" <[email protected]> wrote:
              > >Not really. The JNDI tree replication is done at the cluster-level and
              > >Weblogic doesn't have a concept of a "set" of clusters that communicate.
              > >
              > >What are you trying to accomplish with this approach?
              > >
              > >Peace,
              > >
              > >--
              > >Cameron Purdy
              > >Tangosol, Inc.
              > >Clustering Weblogic? You're either using Coherence, or you should be!
              > >Download a Tangosol Coherence eval today at http://www.tangosol.com/
              > >
              > >
              > >
              > >"Guillaume Bedard" <[email protected]> wrote in message
              > >news:[email protected]...
              > >>
              > >> Hi!
              > >>
              > >> I would like to know if it is possible to have multiple
              > >> clusters of servers (potentially in different network
              > >> zones), and still have a global JNDI tree, as if it
              > >> was only one big cluster? What are the considerations
              > >> to meet such a requirement?
              > >>
              > >> Thanks,
              > >>
              > >> Guillaume Bedard
              > >
              > >
              >
              

Maybe you are looking for

  • Both Internal hard drives fail on Macbook Pro late-2011

    Both internal hard drives (SSD and HDD) failed on Macbook Pro late-2011. Setup: MBP closed with Thunderbolt display attached. Problem on HDD-A I called Apple Support to log an issue with my MBP2011. The system is made up of 2 drives - SSD (128GB) and

  • IPod Touch no longer syncs wirelessly with iTunes

    My 4th gen iPod Touch (IOS 5.1) has been syncing wirelessly with iTunes on my Mac for months without a problem. Suddenly, the iPod no longer appears in the "Devices" area of iTunes, as it always has. "Sync using wireless" is enabled in iTunes iPod To

  • Two sites, Two Exchange servers, same domain

    Exchange can seriously baffle me at the best of times. Which is why I'm writing here at the moment. I have 2 sites in two geographical locations for the same business connected via IPsec VPN. At each site we have: - Domain Controller (domain.local) -

  • PP-PM Integration: Counter based Plan

    Hi All, I am creating a maintenance plan with counter. I have created a measuring point. When I create measuring document the counter gets updated and maintenance plan works fine. When it comes to automatically update the counter when the production

  • Install error itunes 10.3

    Tried to update to iTunes 10.3 and during install I receive an error stating: "install error".  It then continues to finishing installing, but when the installation is done I can not open iTunes, the icon is a default windows icon.