9 Replies Latest reply on Mar 21, 2001 7:41 PM by 3004

    Cannot resolve ClusterAddress: 10.2.227.1:7005,10.2.227.2:7005

    3004
      Hi All,
                
                Various, hopefully well documented, problems occuring. Well I think they
                are problems :-)
                If anybody can shed some light on some, or all, I would appreciate it very
                much.
                
                Trying to set up clustering for the first time...
                I am on a multihomed W2K running WL6 SP1, addresses: 10.2.227.1 and
                10.2.227.2
                I am running admin server on port 7001, and the two other servers on port
                7005, each with its own IP address.
                
                Curious thing: On startup, both servers give a similar message:
                
                <19-Mar-01 17:41:31 GMT> <Info> <Cluster> <Cannot resolve ClusterAddress:
                10.2.227.1:7005,10.2.227.2:7005>
                Starting Cluster Service ....
                <19-Mar-01 17:41:31 GMT> <Info> <Cluster> <Joined cluster MyCluster at
                address 239.200.200.0 on port 7005>
                
                It's not clear to me whether that worked or not. It says it can't resolve
                the cluster address, and then it says its joined the cluster. Is there an
                error there or not? Also, why can't it resolve the clusteraddress? The
                documentation says to list the IP addresses of the machine, which is what I
                have done...
                
                Near the end of the startup sequence, after starting up BOTH, I get in one
                of the servers:
                <19-Mar-01 17:49:18 GMT> <Info> <Cluster> <Adding server
                5618757874964032231S:10.2.227.2:[7005,7005,7002,7002,7005,7002,
                -1] to cluster view>
                <19-Mar-01 17:49:19 GMT> <Info> <Cluster> <Adding
                5618757874964032231S:10.2.227.2:[7005,7005,7002,7002,7005,7002,-1] to
                the cluster>
                
                and in the other server:
                <19-Mar-01 17:49:25 GMT> <Info> <Cluster> <Adding
                server -8904987319081427239S:10.2.227.1:[7005,7005,7002,7002,7005,7002
                ,-1] to cluster view>
                <19-Mar-01 17:49:25 GMT> <Info> <Cluster> <Lost 2 multicast message(s)>
                <19-Mar-01 17:49:28 GMT> <Info> <Cluster>
                <Adding -8904987319081427239S:10.2.227.1:[7005,7005,7002,7002,7005,7002,-1]
                to
                the cluster>
                
                Again, it is not clear whether an error occured. There was the "Lost 2
                multicast messages" but it seems to have added it anyway?
                
                My admin console says both have joined the cluster.
                
                Another thing: When I shut them down, I get something alon gthe lines of:
                
                <19-Mar-01 18:01:32 GMT> <Error> <Cluster> <Disconnecting from cluster
                MyCluster>
                <19-Mar-01 18:01:32 GMT> <Error> <Cluster> <Multicast socket receive error:
                java.net.SocketException: socket closed
                java.net.SocketException: socket closed at
                java.net.PlainDatagramSocketImpl.receive(Native Method) at
                java.net.DatagramSocket.receiveDatagramSocket.java:392) at
                weblogic.cluster.FragmentSocket.receive(FragmentSocket.java:128) at
                weblogic.cluster.MulticastManager.execute(MulticastManager.java:292)
                at weblogic.kernel.ExecuteThread.execute(ExecuteThread.java:137) at
                weblogic.kernel.ExecuteThread.run(ExecuteThread.java:120)>
                <19-Mar-01 18:01:32 GMT> <Info> <Cluster>
                <Removing -8836117622689770913S:10.2.227.2:[7005,7005,7002,7002,7005,7002,-1
                ]
                to the cluster>
                
                Again, its not clear whetehr my cluster was actually working or not. Why
                dod I get a socketexception?
                
                I have run the utils.MulticastTest successfully - well, in that it sends and
                resceives ok:
                eg. E:\bea\wlserver6.0\config\KHDomain>java utils.MulticastTest -N adsf -A
                239.200.200.0 -P 7003
                ***** WARNING ***** WARNING ***** WARNING *****
                Do NOT use the same multicast address as a running WLS cluster.
                Starting test. Hit any key to abort
                Using multicast address 239.200.200.0:7003
                Will send messages under the name adsf every 2 seconds
                Will print warning every 600 seconds if no messages are received
                I (adsf) sent message num 1
                Received message 2 from adsf
                I (adsf) sent message num 2
                ...
                
                I have IIS proxying requests through. MY iisproxy.ini file looks like this:
                #WebLogicHost=10.2.227.1
                #WebLogicPort=7005
                WebLogicCluster=10.2.227.1:7005,10.2.227.2:7005
                ConnectTimeoutSecs=20
                ConnectRetrySecs=2
                Debug=All
                DebugConfigInfo=ON
                
                When I hit 10.2.227.1 it does indeed go through to one of the servers. If I
                take that server down, the requests seem to start going through to the other
                server (then and only then). However, session state is lost. Obviously I
                have to do something else to get session replication to start working...
                
                Curious things happening though:
                a) I get a lot of JSP recompilation - ie. I take server down, bring it up
                again, and it wants to recompile every jsp again!
                b) I get a lot of HTTP 500 errors coming back from IIS. A refresh usually
                gets it to work. See my post in the weblogic.developer.interest.plug-in
                newsgroup for more than this - it appears that IIS is having trouble talking
                to the proxy (which is setup according to the book).
                
                That about ends it! Of course, I have something working - but I doubt
                whether my cluster is fully functional because of all of these errors.
                
                Thanks
                Jon
                
                
                
                
                
                
                
                
        • 1. Re: Cannot resolve ClusterAddress: 10.2.227.1:7005,10.2.227.2:7005
          3004
          This address may be either a DNS host name that maps to multiple IP addresses
                    or a comma separated list of single address host names or IP addresses. It
                    shouldn't contain port #'s.
                    
                    -- Prasad
                    
                    Jon Mountjoy wrote:
                    
                    > Hi All,
                    >
                    > Various, hopefully well documented, problems occuring. Well I think they
                    > are problems :-)
                    > If anybody can shed some light on some, or all, I would appreciate it very
                    > much.
                    >
                    > Trying to set up clustering for the first time...
                    > I am on a multihomed W2K running WL6 SP1, addresses: 10.2.227.1 and
                    > 10.2.227.2
                    > I am running admin server on port 7001, and the two other servers on port
                    > 7005, each with its own IP address.
                    >
                    > Curious thing: On startup, both servers give a similar message:
                    >
                    > <19-Mar-01 17:41:31 GMT> <Info> <Cluster> <Cannot resolve ClusterAddress:
                    > 10.2.227.1:7005,10.2.227.2:7005>
                    > Starting Cluster Service ....
                    > <19-Mar-01 17:41:31 GMT> <Info> <Cluster> <Joined cluster MyCluster at
                    > address 239.200.200.0 on port 7005>
                    >
                    > It's not clear to me whether that worked or not. It says it can't resolve
                    > the cluster address, and then it says its joined the cluster. Is there an
                    > error there or not? Also, why can't it resolve the clusteraddress? The
                    > documentation says to list the IP addresses of the machine, which is what I
                    > have done...
                    >
                    > Near the end of the startup sequence, after starting up BOTH, I get in one
                    > of the servers:
                    > <19-Mar-01 17:49:18 GMT> <Info> <Cluster> <Adding server
                    > 5618757874964032231S:10.2.227.2:[7005,7005,7002,7002,7005,7002,
                    > -1] to cluster view>
                    > <19-Mar-01 17:49:19 GMT> <Info> <Cluster> <Adding
                    > 5618757874964032231S:10.2.227.2:[7005,7005,7002,7002,7005,7002,-1] to
                    > the cluster>
                    >
                    > and in the other server:
                    > <19-Mar-01 17:49:25 GMT> <Info> <Cluster> <Adding
                    > server -8904987319081427239S:10.2.227.1:[7005,7005,7002,7002,7005,7002
                    > ,-1] to cluster view>
                    > <19-Mar-01 17:49:25 GMT> <Info> <Cluster> <Lost 2 multicast message(s)>
                    > <19-Mar-01 17:49:28 GMT> <Info> <Cluster>
                    > <Adding -8904987319081427239S:10.2.227.1:[7005,7005,7002,7002,7005,7002,-1]
                    > to
                    > the cluster>
                    >
                    > Again, it is not clear whether an error occured. There was the "Lost 2
                    > multicast messages" but it seems to have added it anyway?
                    >
                    > My admin console says both have joined the cluster.
                    >
                    > Another thing: When I shut them down, I get something alon gthe lines of:
                    >
                    > <19-Mar-01 18:01:32 GMT> <Error> <Cluster> <Disconnecting from cluster
                    > MyCluster>
                    > <19-Mar-01 18:01:32 GMT> <Error> <Cluster> <Multicast socket receive error:
                    > java.net.SocketException: socket closed
                    > java.net.SocketException: socket closed at
                    > java.net.PlainDatagramSocketImpl.receive(Native Method) at
                    > java.net.DatagramSocket.receiveDatagramSocket.java:392) at
                    > weblogic.cluster.FragmentSocket.receive(FragmentSocket.java:128) at
                    > weblogic.cluster.MulticastManager.execute(MulticastManager.java:292)
                    > at weblogic.kernel.ExecuteThread.execute(ExecuteThread.java:137) at
                    > weblogic.kernel.ExecuteThread.run(ExecuteThread.java:120)>
                    > <19-Mar-01 18:01:32 GMT> <Info> <Cluster>
                    > <Removing -8836117622689770913S:10.2.227.2:[7005,7005,7002,7002,7005,7002,-1
                    > ]
                    > to the cluster>
                    >
                    > Again, its not clear whetehr my cluster was actually working or not. Why
                    > dod I get a socketexception?
                    >
                    > I have run the utils.MulticastTest successfully - well, in that it sends and
                    > resceives ok:
                    > eg. E:\bea\wlserver6.0\config\KHDomain>java utils.MulticastTest -N adsf -A
                    > 239.200.200.0 -P 7003
                    > ***** WARNING ***** WARNING ***** WARNING *****
                    > Do NOT use the same multicast address as a running WLS cluster.
                    > Starting test. Hit any key to abort
                    > Using multicast address 239.200.200.0:7003
                    > Will send messages under the name adsf every 2 seconds
                    > Will print warning every 600 seconds if no messages are received
                    > I (adsf) sent message num 1
                    > Received message 2 from adsf
                    > I (adsf) sent message num 2
                    > ...
                    >
                    > I have IIS proxying requests through. MY iisproxy.ini file looks like this:
                    > #WebLogicHost=10.2.227.1
                    > #WebLogicPort=7005
                    > WebLogicCluster=10.2.227.1:7005,10.2.227.2:7005
                    > ConnectTimeoutSecs=20
                    > ConnectRetrySecs=2
                    > Debug=All
                    > DebugConfigInfo=ON
                    >
                    > When I hit 10.2.227.1 it does indeed go through to one of the servers. If I
                    > take that server down, the requests seem to start going through to the other
                    > server (then and only then). However, session state is lost. Obviously I
                    > have to do something else to get session replication to start working...
                    >
                    > Curious things happening though:
                    > a) I get a lot of JSP recompilation - ie. I take server down, bring it up
                    > again, and it wants to recompile every jsp again!
                    > b) I get a lot of HTTP 500 errors coming back from IIS. A refresh usually
                    > gets it to work. See my post in the weblogic.developer.interest.plug-in
                    > newsgroup for more than this - it appears that IIS is having trouble talking
                    > to the proxy (which is setup according to the book).
                    >
                    > That about ends it! Of course, I have something working - but I doubt
                    > whether my cluster is fully functional because of all of these errors.
                    >
                    > Thanks
                    > Jon
                    
                    
          • 2. Re: Cannot resolve ClusterAddress: 10.2.227.1:7005,10.2.227.2:7005
            3004
            This address may be either a DNS host name that maps to multiple IP
                      addresses
                      > or a comma separated list of single address host names or IP addresses. It
                      > shouldn't contain port #'s.
                      
                      Hello Prasad,
                      
                      I believe I tried it with no port numbers as well, but the other behaviour I
                      describe remains unchanged.
                      
                      Jon
                      
                      
                      
            • 3. Re: Cannot resolve ClusterAddress: 10.2.227.1:7005,10.2.227.2:7005
              3004
              See my comments in line.
                        
                        Jon Mountjoy wrote:
                        
                        > Hi All,
                        >
                        > Various, hopefully well documented, problems occuring. Well I think they
                        > are problems :-)
                        > If anybody can shed some light on some, or all, I would appreciate it very
                        > much.
                        >
                        > Trying to set up clustering for the first time...
                        > I am on a multihomed W2K running WL6 SP1, addresses: 10.2.227.1 and
                        > 10.2.227.2
                        > I am running admin server on port 7001, and the two other servers on port
                        > 7005, each with its own IP address.
                        >
                        > Curious thing: On startup, both servers give a similar message:
                        >
                        > <19-Mar-01 17:41:31 GMT> <Info> <Cluster> <Cannot resolve ClusterAddress:
                        > 10.2.227.1:7005,10.2.227.2:7005>
                        > Starting Cluster Service ....
                        > <19-Mar-01 17:41:31 GMT> <Info> <Cluster> <Joined cluster MyCluster at
                        > address 239.200.200.0 on port 7005>
                        >
                        > It's not clear to me whether that worked or not. It says it can't resolve
                        > the cluster address, and then it says its joined the cluster. Is there an
                        > error there or not? Also, why can't it resolve the clusteraddress? The
                        > documentation says to list the IP addresses of the machine, which is what I
                        > have done...
                        >
                        >
                        
                        We recommend cluster address to be a dns name that can map to multiple ip
                        addresses. You will be fine if you do that. Also, for your initial testing you
                        don't need that property.
                        
                        
                        > Near the end of the startup sequence, after starting up BOTH, I get in one
                        > of the servers:
                        > <19-Mar-01 17:49:18 GMT> <Info> <Cluster> <Adding server
                        > 5618757874964032231S:10.2.227.2:[7005,7005,7002,7002,7005,7002,
                        > -1] to cluster view>
                        > <19-Mar-01 17:49:19 GMT> <Info> <Cluster> <Adding
                        > 5618757874964032231S:10.2.227.2:[7005,7005,7002,7002,7005,7002,-1] to
                        > the cluster>
                        >
                        > and in the other server:
                        > <19-Mar-01 17:49:25 GMT> <Info> <Cluster> <Adding
                        > server -8904987319081427239S:10.2.227.1:[7005,7005,7002,7002,7005,7002
                        > ,-1] to cluster view>
                        > <19-Mar-01 17:49:25 GMT> <Info> <Cluster> <Lost 2 multicast message(s)>
                        > <19-Mar-01 17:49:28 GMT> <Info> <Cluster>
                        > <Adding -8904987319081427239S:10.2.227.1:[7005,7005,7002,7002,7005,7002,-1]
                        > to
                        > the cluster>
                        >
                        > Again, it is not clear whether an error occured. There was the "Lost 2
                        > multicast messages" but it seems to have added it anyway?
                        
                        Multicast is not reliable. Also, by the time you start 2 nd server first
                        server might have already sent messages. Its normal to lose some messages. The
                        only time you have to worry about this is when the cluster stabilize which
                        should be few hundred seconds make sure that you don't keep seeing this.
                        
                        If you see the Adding message in the log, your cluster is fine. Read the
                        replication documentation to solve your other problem.
                        
                        
                        
                        >
                        > My admin console says both have joined the cluster.
                        >
                        > Another thing: When I shut them down, I get something alon gthe lines of:
                        >
                        > <19-Mar-01 18:01:32 GMT> <Error> <Cluster> <Disconnecting from cluster
                        > MyCluster>
                        > <19-Mar-01 18:01:32 GMT> <Error> <Cluster> <Multicast socket receive error:
                        > java.net.SocketException: socket closed
                        > java.net.SocketException: socket closed at
                        > java.net.PlainDatagramSocketImpl.receive(Native Method) at
                        > java.net.DatagramSocket.receiveDatagramSocket.java:392) at
                        > weblogic.cluster.FragmentSocket.receive(FragmentSocket.java:128) at
                        > weblogic.cluster.MulticastManager.execute(MulticastManager.java:292)
                        > at weblogic.kernel.ExecuteThread.execute(ExecuteThread.java:137) at
                        > weblogic.kernel.ExecuteThread.run(ExecuteThread.java:120)>
                        > <19-Mar-01 18:01:32 GMT> <Info> <Cluster>
                        > <Removing -8836117622689770913S:10.2.227.2:[7005,7005,7002,7002,7005,7002,-1
                        > ]
                        > to the cluster>
                        >
                        > Again, its not clear whetehr my cluster was actually working or not. Why
                        > dod I get a socketexception?
                        
                        If you keep seeing this over and over again, then it is a problem if not I won't
                        worry about it. BTW, do you more than one server running on the same machine.
                        
                        >
                        > I have run the utils.MulticastTest successfully - well, in that it sends and
                        > resceives ok:
                        > eg. E:\bea\wlserver6.0\config\KHDomain>java utils.MulticastTest -N adsf -A
                        > 239.200.200.0 -P 7003
                        > ***** WARNING ***** WARNING ***** WARNING *****
                        > Do NOT use the same multicast address as a running WLS cluster.
                        > Starting test. Hit any key to abort
                        > Using multicast address 239.200.200.0:7003
                        > Will send messages under the name adsf every 2 seconds
                        > Will print warning every 600 seconds if no messages are received
                        > I (adsf) sent message num 1
                        > Received message 2 from adsf
                        > I (adsf) sent message num 2
                        > ...
                        >
                        > I have IIS proxying requests through. MY iisproxy.ini file looks like this:
                        > #WebLogicHost=10.2.227.1
                        > #WebLogicPort=7005
                        > WebLogicCluster=10.2.227.1:7005,10.2.227.2:7005
                        > ConnectTimeoutSecs=20
                        > ConnectRetrySecs=2
                        > Debug=All
                        > DebugConfigInfo=ON
                        >
                        > When I hit 10.2.227.1 it does indeed go through to one of the servers. If I
                        > take that server down, the requests seem to start going through to the other
                        > server (then and only then). However, session state is lost. Obviously I
                        > have to do something else to get session replication to start working...
                        >
                        > Curious things happening though:
                        > a) I get a lot of JSP recompilation - ie. I take server down, bring it up
                        > again, and it wants to recompile every jsp again!
                        > b) I get a lot of HTTP 500 errors coming back from IIS. A refresh usually
                        > gets it to work. See my post in the weblogic.developer.interest.plug-in
                        > newsgroup for more than this - it appears that IIS is having trouble talking
                        > to the proxy (which is setup according to the book).
                        >
                        > That about ends it! Of course, I have something working - but I doubt
                        > whether my cluster is fully functional because of all of these errors.
                        >
                        > Thanks
                        > Jon
                        
                        Did you set up your web app to use In Memory replication, if not, you won't get
                        session failover.
                        
                        
                        
                        
              • 4. Re: Cannot resolve ClusterAddress: 10.2.227.1:7005,10.2.227.2:7005
                3004
                Thank you for your comments Prasad.
                          
                          To answer your question - yes, I am running all three servers on one
                          physical (multihomed) machine.
                          
                          
                          
                • 5. Re: Cannot resolve ClusterAddress: 10.2.227.1:7005,10.2.227.2:7005
                  3004
                  I don't know for sure but I am speculating that W2K might be doing
                            something that is causing SocketException. Again I don't know for sure,
                            but as I said before if see this message only once when a server dies I
                            won't worry about this.
                            
                            Cheers,
                            
                            -- Prasad
                            
                            Jon Mountjoy wrote:
                            
                            > Thank you for your comments Prasad.
                            >
                            > To answer your question - yes, I am running all three servers on one
                            > physical (multihomed) machine.
                            
                            
                  • 6. Re: Cannot resolve ClusterAddress: 10.2.227.1:7005,10.2.227.2:7005
                    3004
                    Hi Prasad,
                              
                              What IS the purpose of this ClusterAddress field in 6.0? Is it simply an MBean property from which
                              a client can retrieve the info so it'll know how to connect to the cluster? I'm also getting this
                              "Cannot resolve ClusterAddress" warning, but my ejbs are clustering just fine.
                              
                              Gene
                              
                              "Prasad Peddada" <prasadp@bgnet.bgsu.edu> wrote in message news:3AB6663D.422B7D69@bgnet.bgsu.edu...
                              > This address may be either a DNS host name that maps to multiple IP addresses
                              > or a comma separated list of single address host names or IP addresses. It
                              > shouldn't contain port #'s.
                              >
                              > -- Prasad
                              >
                              > Jon Mountjoy wrote:
                              >
                              > > Hi All,
                              > >
                              > > Various, hopefully well documented, problems occuring. Well I think they
                              > > are problems :-)
                              > > If anybody can shed some light on some, or all, I would appreciate it very
                              > > much.
                              > >
                              > > Trying to set up clustering for the first time...
                              > > I am on a multihomed W2K running WL6 SP1, addresses: 10.2.227.1 and
                              > > 10.2.227.2
                              > > I am running admin server on port 7001, and the two other servers on port
                              > > 7005, each with its own IP address.
                              > >
                              > > Curious thing: On startup, both servers give a similar message:
                              > >
                              > > <19-Mar-01 17:41:31 GMT> <Info> <Cluster> <Cannot resolve ClusterAddress:
                              > > 10.2.227.1:7005,10.2.227.2:7005>
                              > > Starting Cluster Service ....
                              > > <19-Mar-01 17:41:31 GMT> <Info> <Cluster> <Joined cluster MyCluster at
                              > > address 239.200.200.0 on port 7005>
                              > >
                              > > It's not clear to me whether that worked or not. It says it can't resolve
                              > > the cluster address, and then it says its joined the cluster. Is there an
                              > > error there or not? Also, why can't it resolve the clusteraddress? The
                              > > documentation says to list the IP addresses of the machine, which is what I
                              > > have done...
                              > >
                              > > Near the end of the startup sequence, after starting up BOTH, I get in one
                              > > of the servers:
                              > > <19-Mar-01 17:49:18 GMT> <Info> <Cluster> <Adding server
                              > > 5618757874964032231S:10.2.227.2:[7005,7005,7002,7002,7005,7002,
                              > > -1] to cluster view>
                              > > <19-Mar-01 17:49:19 GMT> <Info> <Cluster> <Adding
                              > > 5618757874964032231S:10.2.227.2:[7005,7005,7002,7002,7005,7002,-1] to
                              > > the cluster>
                              > >
                              > > and in the other server:
                              > > <19-Mar-01 17:49:25 GMT> <Info> <Cluster> <Adding
                              > > server -8904987319081427239S:10.2.227.1:[7005,7005,7002,7002,7005,7002
                              > > ,-1] to cluster view>
                              > > <19-Mar-01 17:49:25 GMT> <Info> <Cluster> <Lost 2 multicast message(s)>
                              > > <19-Mar-01 17:49:28 GMT> <Info> <Cluster>
                              > > <Adding -8904987319081427239S:10.2.227.1:[7005,7005,7002,7002,7005,7002,-1]
                              > > to
                              > > the cluster>
                              > >
                              > > Again, it is not clear whether an error occured. There was the "Lost 2
                              > > multicast messages" but it seems to have added it anyway?
                              > >
                              > > My admin console says both have joined the cluster.
                              > >
                              > > Another thing: When I shut them down, I get something alon gthe lines of:
                              > >
                              > > <19-Mar-01 18:01:32 GMT> <Error> <Cluster> <Disconnecting from cluster
                              > > MyCluster>
                              > > <19-Mar-01 18:01:32 GMT> <Error> <Cluster> <Multicast socket receive error:
                              > > java.net.SocketException: socket closed
                              > > java.net.SocketException: socket closed at
                              > > java.net.PlainDatagramSocketImpl.receive(Native Method) at
                              > > java.net.DatagramSocket.receiveDatagramSocket.java:392) at
                              > > weblogic.cluster.FragmentSocket.receive(FragmentSocket.java:128) at
                              > > weblogic.cluster.MulticastManager.execute(MulticastManager.java:292)
                              > > at weblogic.kernel.ExecuteThread.execute(ExecuteThread.java:137) at
                              > > weblogic.kernel.ExecuteThread.run(ExecuteThread.java:120)>
                              > > <19-Mar-01 18:01:32 GMT> <Info> <Cluster>
                              > > <Removing -8836117622689770913S:10.2.227.2:[7005,7005,7002,7002,7005,7002,-1
                              > > ]
                              > > to the cluster>
                              > >
                              > > Again, its not clear whetehr my cluster was actually working or not. Why
                              > > dod I get a socketexception?
                              > >
                              > > I have run the utils.MulticastTest successfully - well, in that it sends and
                              > > resceives ok:
                              > > eg. E:\bea\wlserver6.0\config\KHDomain>java utils.MulticastTest -N adsf -A
                              > > 239.200.200.0 -P 7003
                              > > ***** WARNING ***** WARNING ***** WARNING *****
                              > > Do NOT use the same multicast address as a running WLS cluster.
                              > > Starting test. Hit any key to abort
                              > > Using multicast address 239.200.200.0:7003
                              > > Will send messages under the name adsf every 2 seconds
                              > > Will print warning every 600 seconds if no messages are received
                              > > I (adsf) sent message num 1
                              > > Received message 2 from adsf
                              > > I (adsf) sent message num 2
                              > > ...
                              > >
                              > > I have IIS proxying requests through. MY iisproxy.ini file looks like this:
                              > > #WebLogicHost=10.2.227.1
                              > > #WebLogicPort=7005
                              > > WebLogicCluster=10.2.227.1:7005,10.2.227.2:7005
                              > > ConnectTimeoutSecs=20
                              > > ConnectRetrySecs=2
                              > > Debug=All
                              > > DebugConfigInfo=ON
                              > >
                              > > When I hit 10.2.227.1 it does indeed go through to one of the servers. If I
                              > > take that server down, the requests seem to start going through to the other
                              > > server (then and only then). However, session state is lost. Obviously I
                              > > have to do something else to get session replication to start working...
                              > >
                              > > Curious things happening though:
                              > > a) I get a lot of JSP recompilation - ie. I take server down, bring it up
                              > > again, and it wants to recompile every jsp again!
                              > > b) I get a lot of HTTP 500 errors coming back from IIS. A refresh usually
                              > > gets it to work. See my post in the weblogic.developer.interest.plug-in
                              > > newsgroup for more than this - it appears that IIS is having trouble talking
                              > > to the proxy (which is setup according to the book).
                              > >
                              > > That about ends it! Of course, I have something working - but I doubt
                              > > whether my cluster is fully functional because of all of these errors.
                              > >
                              > > Thanks
                              > > Jon
                              >
                              
                              
                              
                    • 7. Re: Cannot resolve ClusterAddress: 10.2.227.1:7005,10.2.227.2:7005
                      3004
                      If you are using handles you have to set this property. You don't need otherwise for current release.
                                
                                Cheers,
                                
                                -- Prasad
                                
                                
                                Gene Chuang wrote:
                                
                                > Hi Prasad,
                                >
                                > What IS the purpose of this ClusterAddress field in 6.0? Is it simply an MBean property from which
                                > a client can retrieve the info so it'll know how to connect to the cluster? I'm also getting this
                                > "Cannot resolve ClusterAddress" warning, but my ejbs are clustering just fine.
                                >
                                > Gene
                                >
                                > "Prasad Peddada" <prasadp@bgnet.bgsu.edu> wrote in message news:3AB6663D.422B7D69@bgnet.bgsu.edu...
                                > > This address may be either a DNS host name that maps to multiple IP addresses
                                > > or a comma separated list of single address host names or IP addresses. It
                                > > shouldn't contain port #'s.
                                > >
                                > > -- Prasad
                                > >
                                > > Jon Mountjoy wrote:
                                > >
                                > > > Hi All,
                                > > >
                                > > > Various, hopefully well documented, problems occuring. Well I think they
                                > > > are problems :-)
                                > > > If anybody can shed some light on some, or all, I would appreciate it very
                                > > > much.
                                > > >
                                > > > Trying to set up clustering for the first time...
                                > > > I am on a multihomed W2K running WL6 SP1, addresses: 10.2.227.1 and
                                > > > 10.2.227.2
                                > > > I am running admin server on port 7001, and the two other servers on port
                                > > > 7005, each with its own IP address.
                                > > >
                                > > > Curious thing: On startup, both servers give a similar message:
                                > > >
                                > > > <19-Mar-01 17:41:31 GMT> <Info> <Cluster> <Cannot resolve ClusterAddress:
                                > > > 10.2.227.1:7005,10.2.227.2:7005>
                                > > > Starting Cluster Service ....
                                > > > <19-Mar-01 17:41:31 GMT> <Info> <Cluster> <Joined cluster MyCluster at
                                > > > address 239.200.200.0 on port 7005>
                                > > >
                                > > > It's not clear to me whether that worked or not. It says it can't resolve
                                > > > the cluster address, and then it says its joined the cluster. Is there an
                                > > > error there or not? Also, why can't it resolve the clusteraddress? The
                                > > > documentation says to list the IP addresses of the machine, which is what I
                                > > > have done...
                                > > >
                                > > > Near the end of the startup sequence, after starting up BOTH, I get in one
                                > > > of the servers:
                                > > > <19-Mar-01 17:49:18 GMT> <Info> <Cluster> <Adding server
                                > > > 5618757874964032231S:10.2.227.2:[7005,7005,7002,7002,7005,7002,
                                > > > -1] to cluster view>
                                > > > <19-Mar-01 17:49:19 GMT> <Info> <Cluster> <Adding
                                > > > 5618757874964032231S:10.2.227.2:[7005,7005,7002,7002,7005,7002,-1] to
                                > > > the cluster>
                                > > >
                                > > > and in the other server:
                                > > > <19-Mar-01 17:49:25 GMT> <Info> <Cluster> <Adding
                                > > > server -8904987319081427239S:10.2.227.1:[7005,7005,7002,7002,7005,7002
                                > > > ,-1] to cluster view>
                                > > > <19-Mar-01 17:49:25 GMT> <Info> <Cluster> <Lost 2 multicast message(s)>
                                > > > <19-Mar-01 17:49:28 GMT> <Info> <Cluster>
                                > > > <Adding -8904987319081427239S:10.2.227.1:[7005,7005,7002,7002,7005,7002,-1]
                                > > > to
                                > > > the cluster>
                                > > >
                                > > > Again, it is not clear whether an error occured. There was the "Lost 2
                                > > > multicast messages" but it seems to have added it anyway?
                                > > >
                                > > > My admin console says both have joined the cluster.
                                > > >
                                > > > Another thing: When I shut them down, I get something alon gthe lines of:
                                > > >
                                > > > <19-Mar-01 18:01:32 GMT> <Error> <Cluster> <Disconnecting from cluster
                                > > > MyCluster>
                                > > > <19-Mar-01 18:01:32 GMT> <Error> <Cluster> <Multicast socket receive error:
                                > > > java.net.SocketException: socket closed
                                > > > java.net.SocketException: socket closed at
                                > > > java.net.PlainDatagramSocketImpl.receive(Native Method) at
                                > > > java.net.DatagramSocket.receiveDatagramSocket.java:392) at
                                > > > weblogic.cluster.FragmentSocket.receive(FragmentSocket.java:128) at
                                > > > weblogic.cluster.MulticastManager.execute(MulticastManager.java:292)
                                > > > at weblogic.kernel.ExecuteThread.execute(ExecuteThread.java:137) at
                                > > > weblogic.kernel.ExecuteThread.run(ExecuteThread.java:120)>
                                > > > <19-Mar-01 18:01:32 GMT> <Info> <Cluster>
                                > > > <Removing -8836117622689770913S:10.2.227.2:[7005,7005,7002,7002,7005,7002,-1
                                > > > ]
                                > > > to the cluster>
                                > > >
                                > > > Again, its not clear whetehr my cluster was actually working or not. Why
                                > > > dod I get a socketexception?
                                > > >
                                > > > I have run the utils.MulticastTest successfully - well, in that it sends and
                                > > > resceives ok:
                                > > > eg. E:\bea\wlserver6.0\config\KHDomain>java utils.MulticastTest -N adsf -A
                                > > > 239.200.200.0 -P 7003
                                > > > ***** WARNING ***** WARNING ***** WARNING *****
                                > > > Do NOT use the same multicast address as a running WLS cluster.
                                > > > Starting test. Hit any key to abort
                                > > > Using multicast address 239.200.200.0:7003
                                > > > Will send messages under the name adsf every 2 seconds
                                > > > Will print warning every 600 seconds if no messages are received
                                > > > I (adsf) sent message num 1
                                > > > Received message 2 from adsf
                                > > > I (adsf) sent message num 2
                                > > > ...
                                > > >
                                > > > I have IIS proxying requests through. MY iisproxy.ini file looks like this:
                                > > > #WebLogicHost=10.2.227.1
                                > > > #WebLogicPort=7005
                                > > > WebLogicCluster=10.2.227.1:7005,10.2.227.2:7005
                                > > > ConnectTimeoutSecs=20
                                > > > ConnectRetrySecs=2
                                > > > Debug=All
                                > > > DebugConfigInfo=ON
                                > > >
                                > > > When I hit 10.2.227.1 it does indeed go through to one of the servers. If I
                                > > > take that server down, the requests seem to start going through to the other
                                > > > server (then and only then). However, session state is lost. Obviously I
                                > > > have to do something else to get session replication to start working...
                                > > >
                                > > > Curious things happening though:
                                > > > a) I get a lot of JSP recompilation - ie. I take server down, bring it up
                                > > > again, and it wants to recompile every jsp again!
                                > > > b) I get a lot of HTTP 500 errors coming back from IIS. A refresh usually
                                > > > gets it to work. See my post in the weblogic.developer.interest.plug-in
                                > > > newsgroup for more than this - it appears that IIS is having trouble talking
                                > > > to the proxy (which is setup according to the book).
                                > > >
                                > > > That about ends it! Of course, I have something working - but I doubt
                                > > > whether my cluster is fully functional because of all of these errors.
                                > > >
                                > > > Thanks
                                > > > Jon
                                > >
                                
                                
                      • 8. Re: Cannot resolve ClusterAddress: 10.2.227.1:7005,10.2.227.2:7005
                        3004
                        I see: handles use it re-lookup the references. So this property is used only by the container and
                                  cannot be retrieved elsewhere?
                                  
                                  Gene
                                  
                                  "Prasad Peddada" <prasadp@bgnet.bgsu.edu> wrote in message news:3AB81D30.BAFE7181@bgnet.bgsu.edu...
                                  > If you are using handles you have to set this property. You don't need otherwise for current
                                  release.
                                  >
                                  > Cheers,
                                  >
                                  > -- Prasad
                                  >
                                  >
                                  > Gene Chuang wrote:
                                  >
                                  > > Hi Prasad,
                                  > >
                                  > > What IS the purpose of this ClusterAddress field in 6.0? Is it simply an MBean property from
                                  which
                                  > > a client can retrieve the info so it'll know how to connect to the cluster? I'm also getting
                                  this
                                  > > "Cannot resolve ClusterAddress" warning, but my ejbs are clustering just fine.
                                  > >
                                  > > Gene
                                  > >
                                  > > "Prasad Peddada" <prasadp@bgnet.bgsu.edu> wrote in message
                                  news:3AB6663D.422B7D69@bgnet.bgsu.edu...
                                  > > > This address may be either a DNS host name that maps to multiple IP addresses
                                  > > > or a comma separated list of single address host names or IP addresses. It
                                  > > > shouldn't contain port #'s.
                                  > > >
                                  > > > -- Prasad
                                  > > >
                                  > > > Jon Mountjoy wrote:
                                  > > >
                                  > > > > Hi All,
                                  > > > >
                                  > > > > Various, hopefully well documented, problems occuring. Well I think they
                                  > > > > are problems :-)
                                  > > > > If anybody can shed some light on some, or all, I would appreciate it very
                                  > > > > much.
                                  > > > >
                                  > > > > Trying to set up clustering for the first time...
                                  > > > > I am on a multihomed W2K running WL6 SP1, addresses: 10.2.227.1 and
                                  > > > > 10.2.227.2
                                  > > > > I am running admin server on port 7001, and the two other servers on port
                                  > > > > 7005, each with its own IP address.
                                  > > > >
                                  > > > > Curious thing: On startup, both servers give a similar message:
                                  > > > >
                                  > > > > <19-Mar-01 17:41:31 GMT> <Info> <Cluster> <Cannot resolve ClusterAddress:
                                  > > > > 10.2.227.1:7005,10.2.227.2:7005>
                                  > > > > Starting Cluster Service ....
                                  > > > > <19-Mar-01 17:41:31 GMT> <Info> <Cluster> <Joined cluster MyCluster at
                                  > > > > address 239.200.200.0 on port 7005>
                                  > > > >
                                  > > > > It's not clear to me whether that worked or not. It says it can't resolve
                                  > > > > the cluster address, and then it says its joined the cluster. Is there an
                                  > > > > error there or not? Also, why can't it resolve the clusteraddress? The
                                  > > > > documentation says to list the IP addresses of the machine, which is what I
                                  > > > > have done...
                                  > > > >
                                  > > > > Near the end of the startup sequence, after starting up BOTH, I get in one
                                  > > > > of the servers:
                                  > > > > <19-Mar-01 17:49:18 GMT> <Info> <Cluster> <Adding server
                                  > > > > 5618757874964032231S:10.2.227.2:[7005,7005,7002,7002,7005,7002,
                                  > > > > -1] to cluster view>
                                  > > > > <19-Mar-01 17:49:19 GMT> <Info> <Cluster> <Adding
                                  > > > > 5618757874964032231S:10.2.227.2:[7005,7005,7002,7002,7005,7002,-1] to
                                  > > > > the cluster>
                                  > > > >
                                  > > > > and in the other server:
                                  > > > > <19-Mar-01 17:49:25 GMT> <Info> <Cluster> <Adding
                                  > > > > server -8904987319081427239S:10.2.227.1:[7005,7005,7002,7002,7005,7002
                                  > > > > ,-1] to cluster view>
                                  > > > > <19-Mar-01 17:49:25 GMT> <Info> <Cluster> <Lost 2 multicast message(s)>
                                  > > > > <19-Mar-01 17:49:28 GMT> <Info> <Cluster>
                                  > > > > <Adding -8904987319081427239S:10.2.227.1:[7005,7005,7002,7002,7005,7002,-1]
                                  > > > > to
                                  > > > > the cluster>
                                  > > > >
                                  > > > > Again, it is not clear whether an error occured. There was the "Lost 2
                                  > > > > multicast messages" but it seems to have added it anyway?
                                  > > > >
                                  > > > > My admin console says both have joined the cluster.
                                  > > > >
                                  > > > > Another thing: When I shut them down, I get something alon gthe lines of:
                                  > > > >
                                  > > > > <19-Mar-01 18:01:32 GMT> <Error> <Cluster> <Disconnecting from cluster
                                  > > > > MyCluster>
                                  > > > > <19-Mar-01 18:01:32 GMT> <Error> <Cluster> <Multicast socket receive error:
                                  > > > > java.net.SocketException: socket closed
                                  > > > > java.net.SocketException: socket closed at
                                  > > > > java.net.PlainDatagramSocketImpl.receive(Native Method) at
                                  > > > > java.net.DatagramSocket.receiveDatagramSocket.java:392) at
                                  > > > > weblogic.cluster.FragmentSocket.receive(FragmentSocket.java:128) at
                                  > > > > weblogic.cluster.MulticastManager.execute(MulticastManager.java:292)
                                  > > > > at weblogic.kernel.ExecuteThread.execute(ExecuteThread.java:137) at
                                  > > > > weblogic.kernel.ExecuteThread.run(ExecuteThread.java:120)>
                                  > > > > <19-Mar-01 18:01:32 GMT> <Info> <Cluster>
                                  > > > > <Removing -8836117622689770913S:10.2.227.2:[7005,7005,7002,7002,7005,7002,-1
                                  > > > > ]
                                  > > > > to the cluster>
                                  > > > >
                                  > > > > Again, its not clear whetehr my cluster was actually working or not. Why
                                  > > > > dod I get a socketexception?
                                  > > > >
                                  > > > > I have run the utils.MulticastTest successfully - well, in that it sends and
                                  > > > > resceives ok:
                                  > > > > eg. E:\bea\wlserver6.0\config\KHDomain>java utils.MulticastTest -N adsf -A
                                  > > > > 239.200.200.0 -P 7003
                                  > > > > ***** WARNING ***** WARNING ***** WARNING *****
                                  > > > > Do NOT use the same multicast address as a running WLS cluster.
                                  > > > > Starting test. Hit any key to abort
                                  > > > > Using multicast address 239.200.200.0:7003
                                  > > > > Will send messages under the name adsf every 2 seconds
                                  > > > > Will print warning every 600 seconds if no messages are received
                                  > > > > I (adsf) sent message num 1
                                  > > > > Received message 2 from adsf
                                  > > > > I (adsf) sent message num 2
                                  > > > > ...
                                  > > > >
                                  > > > > I have IIS proxying requests through. MY iisproxy.ini file looks like this:
                                  > > > > #WebLogicHost=10.2.227.1
                                  > > > > #WebLogicPort=7005
                                  > > > > WebLogicCluster=10.2.227.1:7005,10.2.227.2:7005
                                  > > > > ConnectTimeoutSecs=20
                                  > > > > ConnectRetrySecs=2
                                  > > > > Debug=All
                                  > > > > DebugConfigInfo=ON
                                  > > > >
                                  > > > > When I hit 10.2.227.1 it does indeed go through to one of the servers. If I
                                  > > > > take that server down, the requests seem to start going through to the other
                                  > > > > server (then and only then). However, session state is lost. Obviously I
                                  > > > > have to do something else to get session replication to start working...
                                  > > > >
                                  > > > > Curious things happening though:
                                  > > > > a) I get a lot of JSP recompilation - ie. I take server down, bring it up
                                  > > > > again, and it wants to recompile every jsp again!
                                  > > > > b) I get a lot of HTTP 500 errors coming back from IIS. A refresh usually
                                  > > > > gets it to work. See my post in the weblogic.developer.interest.plug-in
                                  > > > > newsgroup for more than this - it appears that IIS is having trouble talking
                                  > > > > to the proxy (which is setup according to the book).
                                  > > > >
                                  > > > > That about ends it! Of course, I have something working - but I doubt
                                  > > > > whether my cluster is fully functional because of all of these errors.
                                  > > > >
                                  > > > > Thanks
                                  > > > > Jon
                                  > > >
                                  >
                                  
                                  
                                  
                        • 9. Re: Cannot resolve ClusterAddress: 10.2.227.1:7005,10.2.227.2:7005
                          3004
                          yep.
                                    
                                    -- prasad
                                    
                                    Gene Chuang wrote:
                                    
                                    > I see: handles use it re-lookup the references. So this property is used only by the container and
                                    > cannot be retrieved elsewhere?
                                    >
                                    > Gene
                                    >
                                    > "Prasad Peddada" <prasadp@bgnet.bgsu.edu> wrote in message news:3AB81D30.BAFE7181@bgnet.bgsu.edu...
                                    > > If you are using handles you have to set this property. You don't need otherwise for current
                                    > release.
                                    > >
                                    > > Cheers,
                                    > >
                                    > > -- Prasad
                                    > >
                                    > >
                                    > > Gene Chuang wrote:
                                    > >
                                    > > > Hi Prasad,
                                    > > >
                                    > > > What IS the purpose of this ClusterAddress field in 6.0? Is it simply an MBean property from
                                    > which
                                    > > > a client can retrieve the info so it'll know how to connect to the cluster? I'm also getting
                                    > this
                                    > > > "Cannot resolve ClusterAddress" warning, but my ejbs are clustering just fine.
                                    > > >
                                    > > > Gene
                                    > > >
                                    > > > "Prasad Peddada" <prasadp@bgnet.bgsu.edu> wrote in message
                                    > news:3AB6663D.422B7D69@bgnet.bgsu.edu...
                                    > > > > This address may be either a DNS host name that maps to multiple IP addresses
                                    > > > > or a comma separated list of single address host names or IP addresses. It
                                    > > > > shouldn't contain port #'s.
                                    > > > >
                                    > > > > -- Prasad
                                    > > > >
                                    > > > > Jon Mountjoy wrote:
                                    > > > >
                                    > > > > > Hi All,
                                    > > > > >
                                    > > > > > Various, hopefully well documented, problems occuring. Well I think they
                                    > > > > > are problems :-)
                                    > > > > > If anybody can shed some light on some, or all, I would appreciate it very
                                    > > > > > much.
                                    > > > > >
                                    > > > > > Trying to set up clustering for the first time...
                                    > > > > > I am on a multihomed W2K running WL6 SP1, addresses: 10.2.227.1 and
                                    > > > > > 10.2.227.2
                                    > > > > > I am running admin server on port 7001, and the two other servers on port
                                    > > > > > 7005, each with its own IP address.
                                    > > > > >
                                    > > > > > Curious thing: On startup, both servers give a similar message:
                                    > > > > >
                                    > > > > > <19-Mar-01 17:41:31 GMT> <Info> <Cluster> <Cannot resolve ClusterAddress:
                                    > > > > > 10.2.227.1:7005,10.2.227.2:7005>
                                    > > > > > Starting Cluster Service ....
                                    > > > > > <19-Mar-01 17:41:31 GMT> <Info> <Cluster> <Joined cluster MyCluster at
                                    > > > > > address 239.200.200.0 on port 7005>
                                    > > > > >
                                    > > > > > It's not clear to me whether that worked or not. It says it can't resolve
                                    > > > > > the cluster address, and then it says its joined the cluster. Is there an
                                    > > > > > error there or not? Also, why can't it resolve the clusteraddress? The
                                    > > > > > documentation says to list the IP addresses of the machine, which is what I
                                    > > > > > have done...
                                    > > > > >
                                    > > > > > Near the end of the startup sequence, after starting up BOTH, I get in one
                                    > > > > > of the servers:
                                    > > > > > <19-Mar-01 17:49:18 GMT> <Info> <Cluster> <Adding server
                                    > > > > > 5618757874964032231S:10.2.227.2:[7005,7005,7002,7002,7005,7002,
                                    > > > > > -1] to cluster view>
                                    > > > > > <19-Mar-01 17:49:19 GMT> <Info> <Cluster> <Adding
                                    > > > > > 5618757874964032231S:10.2.227.2:[7005,7005,7002,7002,7005,7002,-1] to
                                    > > > > > the cluster>
                                    > > > > >
                                    > > > > > and in the other server:
                                    > > > > > <19-Mar-01 17:49:25 GMT> <Info> <Cluster> <Adding
                                    > > > > > server -8904987319081427239S:10.2.227.1:[7005,7005,7002,7002,7005,7002
                                    > > > > > ,-1] to cluster view>
                                    > > > > > <19-Mar-01 17:49:25 GMT> <Info> <Cluster> <Lost 2 multicast message(s)>
                                    > > > > > <19-Mar-01 17:49:28 GMT> <Info> <Cluster>
                                    > > > > > <Adding -8904987319081427239S:10.2.227.1:[7005,7005,7002,7002,7005,7002,-1]
                                    > > > > > to
                                    > > > > > the cluster>
                                    > > > > >
                                    > > > > > Again, it is not clear whether an error occured. There was the "Lost 2
                                    > > > > > multicast messages" but it seems to have added it anyway?
                                    > > > > >
                                    > > > > > My admin console says both have joined the cluster.
                                    > > > > >
                                    > > > > > Another thing: When I shut them down, I get something alon gthe lines of:
                                    > > > > >
                                    > > > > > <19-Mar-01 18:01:32 GMT> <Error> <Cluster> <Disconnecting from cluster
                                    > > > > > MyCluster>
                                    > > > > > <19-Mar-01 18:01:32 GMT> <Error> <Cluster> <Multicast socket receive error:
                                    > > > > > java.net.SocketException: socket closed
                                    > > > > > java.net.SocketException: socket closed at
                                    > > > > > java.net.PlainDatagramSocketImpl.receive(Native Method) at
                                    > > > > > java.net.DatagramSocket.receiveDatagramSocket.java:392) at
                                    > > > > > weblogic.cluster.FragmentSocket.receive(FragmentSocket.java:128) at
                                    > > > > > weblogic.cluster.MulticastManager.execute(MulticastManager.java:292)
                                    > > > > > at weblogic.kernel.ExecuteThread.execute(ExecuteThread.java:137) at
                                    > > > > > weblogic.kernel.ExecuteThread.run(ExecuteThread.java:120)>
                                    > > > > > <19-Mar-01 18:01:32 GMT> <Info> <Cluster>
                                    > > > > > <Removing -8836117622689770913S:10.2.227.2:[7005,7005,7002,7002,7005,7002,-1
                                    > > > > > ]
                                    > > > > > to the cluster>
                                    > > > > >
                                    > > > > > Again, its not clear whetehr my cluster was actually working or not. Why
                                    > > > > > dod I get a socketexception?
                                    > > > > >
                                    > > > > > I have run the utils.MulticastTest successfully - well, in that it sends and
                                    > > > > > resceives ok:
                                    > > > > > eg. E:\bea\wlserver6.0\config\KHDomain>java utils.MulticastTest -N adsf -A
                                    > > > > > 239.200.200.0 -P 7003
                                    > > > > > ***** WARNING ***** WARNING ***** WARNING *****
                                    > > > > > Do NOT use the same multicast address as a running WLS cluster.
                                    > > > > > Starting test. Hit any key to abort
                                    > > > > > Using multicast address 239.200.200.0:7003
                                    > > > > > Will send messages under the name adsf every 2 seconds
                                    > > > > > Will print warning every 600 seconds if no messages are received
                                    > > > > > I (adsf) sent message num 1
                                    > > > > > Received message 2 from adsf
                                    > > > > > I (adsf) sent message num 2
                                    > > > > > ...
                                    > > > > >
                                    > > > > > I have IIS proxying requests through. MY iisproxy.ini file looks like this:
                                    > > > > > #WebLogicHost=10.2.227.1
                                    > > > > > #WebLogicPort=7005
                                    > > > > > WebLogicCluster=10.2.227.1:7005,10.2.227.2:7005
                                    > > > > > ConnectTimeoutSecs=20
                                    > > > > > ConnectRetrySecs=2
                                    > > > > > Debug=All
                                    > > > > > DebugConfigInfo=ON
                                    > > > > >
                                    > > > > > When I hit 10.2.227.1 it does indeed go through to one of the servers. If I
                                    > > > > > take that server down, the requests seem to start going through to the other
                                    > > > > > server (then and only then). However, session state is lost. Obviously I
                                    > > > > > have to do something else to get session replication to start working...
                                    > > > > >
                                    > > > > > Curious things happening though:
                                    > > > > > a) I get a lot of JSP recompilation - ie. I take server down, bring it up
                                    > > > > > again, and it wants to recompile every jsp again!
                                    > > > > > b) I get a lot of HTTP 500 errors coming back from IIS. A refresh usually
                                    > > > > > gets it to work. See my post in the weblogic.developer.interest.plug-in
                                    > > > > > newsgroup for more than this - it appears that IIS is having trouble talking
                                    > > > > > to the proxy (which is setup according to the book).
                                    > > > > >
                                    > > > > > That about ends it! Of course, I have something working - but I doubt
                                    > > > > > whether my cluster is fully functional because of all of these errors.
                                    > > > > >
                                    > > > > > Thanks
                                    > > > > > Jon
                                    > > > >
                                    > >
                                    
                                    --
                                    Cheers
                                    
                                    - Prasad