We have two integrations configured to be Clustered with internal load balancing enabled.
But we are having the following exception regularly immediately on the startup of the Integration Server.

[ISS.0033.9998E] Exception –> com.wm.io.comm.CommException: [link pool is empty] link pool is empty

Has anybody faced similar kind of exception.?
Any inputs on this will be of great help.

TIA,
Sai Sekhar

Sai,

Make sure your repo server started up ok. Check the repo logs.
Also it is better to use a external load balancer instead of
using the IS for that purpose, since it will be the single point
of failure.

Hi Mike,

I see everything perfect.
The reposerver is running fine and i donot see any errors in the reposerver log. I do understand your point that it is advised to use external load balancer w.r.t single point of failure. As clustering ensures that if one node is down, then the other is available to provide the service. However, I do understand that there are some issues around load balancing strategy employed internally by Integration Server. But this exception seems to be something strange which I am not able to figure out and this exception is observed just after the successful initialization of the Integration Servers.

To troubleshoot this, I would do the following and posts results on these;

1. write a dummy flow Svc to ping the other server in the cluster; run this Svc on both the servers to ensure that n/w connectivity
2. check the server.ThreadPool config param and see if there is any problem there

Does this exception occur on both IS in the cluster?

If you already have a solution for this, please share it.

This exception happens only once just after the successful initialization of the integration server. There is no more entry thereafter. So, I doubt anything wrong with the n/w.

Ok. Then what is the server.threadPool value versus number of system threads on the server statistics page on your servers?

You can use extended settings or server.cnf file to get these param values for me; “watt.server.threadPool”

Do you know for sure that Clustering/load balancing is working despite this error?

Among other things, did you try increasing the log level to get more information on the error. also, try to check the Audit logs to see what/who is the originator of the error.

Also, what is the load balancing option set to? If it isnt (Round Robin Using Redirection), I would prefer you trying that option and see if such an error occurs?

I am inclining that this error is somehow related/associated to “Cluster Links”?

Please try the above and let me know of the results.

Hope we can resolve this soon.

Did you try to reconfigure the cluster? Disable the cluster on both servers, restart both IS’s and reconfigure the cluster. We had the same issue, but we solved it this way.

Is there any way we can have two webMethods Repo Servers running on the same host but on different ports.?

The objective is two have two instances each on two different hosts and configure clustering between M/c1 of Host 1 & M/c1 of Host2 and also have the clustering configured between M/c2 of Host1 & M/c2 of Host2.

It will be of great help if i can have the inputs and configuration details.

TIA,
Sai Sekhar

have two machines hosting Two separate instances of Integration Server each.

There are two clusters defined across the instances of the integration servers running on the m/cs. The reposerver is running fine without any exceptions. Upon starting the integration servers, following exception is trhown.

2004-07-06 11:40:51 IST [ISC.0067.0108C] initRemoteRepositoryClient() exception: RMI Exception: Naming lookup of Repository failed, Error marshaling transport header; nested exception is:
java.io.EOFException

The integration servers take long time for successful startup and the cluster seems to be non functional.

Any advise will be highly appreciated.

Thanks,
Sai Sekhar

Shekhar,
This error might come when its not able to get the right class files.I was getting the same error but i resolved when I put the right class files.Can you tell which version of IS and Oracle you are using. And check if you have put the jar files in the <integration_server>/lib/system path.This is because Repository server picks up classes from the system subfolder.This thing might solve ur problem.Let me know if it works.

Thanks,
Puneet Verma

Shekhar,
Just a basic question though…have u started ur Repository servers before starting integration server.

Thanks,
Puneet Verma

We are also receiving the “link pool is empty” error.
This error (ISS.0033.9998E) is not mentioned in the IS 6.0.1 “Error Message Reference”. Does anyone know what the message is actually saying? I intend to open a problem with webMethods today.
There are no errors in the reposerver2 log. I can find nothing in the configuration (server.cnf) that is missing/different from other IS clusters that are correctly working.
2004-07-13 06:26:32 [ISS.0033.0109D] Server 158.35.100.65 added to cluster WEBM_CLUSTER_NAME
2004-07-13 06:26:32 [ISS.0080.0101V2] SocketLink successfully created to endpoint:158.35.100.65:8897
2004-07-13 06:26:32 [ISS.0080.0101V2] SocketLink successfully created to endpoint:158.35.100.65:4667
2004-07-13 06:26:32 [ISS.0079.0106V2] Received Socket Registration Request from host:158.35.100.65:4667
2004-07-13 06:26:33 [ISS.0025.0016C] Config File Directory Saved
2004-07-13 06:26:33 [ISS.0014.0002C] Initialization completed in 2212 seconds.
2004-07-13 06:26:39 [ISS.0080.0101V2] SocketLink successfully created to endpoint:158.35.100.11:2947
2004-07-13 06:26:39 [ISS.0079.0106V2] Received Socket Registration Request from host:158.35.100.11:2947
2004-07-13 06:27:02 [ISS.0079.0101V2] Unable to register Reverse Connection with alias:CLUSTER158.35.100.65
2004-07-13 06:27:02 [ISS.0080.0101V2] SocketLink successfully created to endpoint:158.35.100.65:8897
2004-07-13 06:27:02 [ISS.0080.0101V2] SocketLink successfully created to endpoint:158.35.100.65:4670
2004-07-13 06:27:02 [ISS.0079.0106V2] Received Socket Registration Request from host:158.35.100.65:4670
2004-07-13 06:27:10 [ISS.0080.0101V2] SocketLink successfully created to endpoint:158.35.100.11:2951
2004-07-13 06:27:10 [ISS.0079.0106V2] Received Socket Registration Request from host:158.35.100.11:2951
2004-07-13 06:27:52 [ISS.0079.0101V2] Unable to register Reverse Connection with alias:CLUSTER158.35.100.65
2004-07-13 06:27:52 [ISS.0080.0101V2] SocketLink successfully created to endpoint:158.35.100.65:8897
2004-07-13 06:27:52 [ISS.0080.0101V2] SocketLink successfully created to endpoint:158.35.100.65:4681
2004-07-13 06:27:52 [ISS.0079.0106V2] Received Socket Registration Request from host:158.35.100.65:4681
2004-07-13 06:27:55 [ISS.0080.0101V2] SocketLink successfully created to endpoint:158.35.100.11:2957
2004-07-13 06:27:55 [ISS.0079.0106V2] Received Socket Registration Request from host:158.35.100.11:2957
2004-07-13 06:28:22 [ISS.0079.0101V2] Unable to register Reverse Connection with alias:CLUSTER158.35.100.65
2004-07-13 06:28:22 [ISS.0079.0100V2] Registered 0 Reverse Connections with alias:CLUSTER158.35.100.65
2004-07-13 06:28:22 [ISS.0080.0101V2] SocketLink successfully created to endpoint:158.35.100.11:8897
2004-07-13 06:29:11 [ISS.0079.0101V2] Unable to register Reverse Connection with alias:CLUSTER158.35.100.11
2004-07-13 06:29:11 [ISS.0080.0101V2] SocketLink successfully created to endpoint:158.35.100.11:8897
2004-07-13 06:29:11 [ISS.0033.9998E] Exception –> com.wm.io.comm.CommException: [link pool is empty] link pool is empty
2004-07-13 06:29:39 [ISS.0080.0101V2] SocketLink successfully created to endpoint:158.35.100.11:2971
2004-07-13 06:29:39 [ISS.0079.0106V2] Received Socket Registration Request from host:158.35.100.11:2971
2004-07-13 06:29:41 [ISS.0079.0101V2] Unable to register Reverse Connection with alias:CLUSTER158.35.100.11
2004-07-13 06:29:41 [ISS.0080.0101V2] SocketLink successfully created to endpoint:158.35.100.11:8897
2004-07-13 06:30:31 [ISS.0079.0101V2] Unable to register Reverse Connection with alias:CLUSTER158.35.100.11
2004-07-13 06:30:31 [ISS.0079.0100V2] Registered 0 Reverse Connections with alias:CLUSTER158.35.100.11

Hi all,
I am working on clustering(IS 6.1, windows), after creating the repository servers in the admin page it is asking to restart the server
when i am restarting the server the following error occurs ,
could any one help me please…
005-06-30 10:02:50 EDT [ART.0116.3509V2] Adapter Runtime (Notification): Notifi
cation Manager waiting on Integration Server initialization.
2005-06-30 10:02:53 EDT [ART.0116.3509V2] Adapter Runtime (Notification): Notifi
cation Manager waiting on Integration Server initialization.
2005-06-30 10:02:56 EDT [ART.0116.3509V2] Adapter Runtime (Notification): Notifi
cation Manager waiting on Integration Server initialization.
2005-06-30 10:02:59 EDT [ISS.0079.0101V2] Unable to register Reverse Connection
with alias:CLUSTERmirdev72
2005-06-30 10:02:59 EDT [ART.0116.3509V2] Adapter Runtime (Notification): Notifi
cation Manager waiting on Integration Server initialization.
2005-06-30 10:03:00 EDT [ISS.0079.0101V2] Unable to register Reverse Connection
with alias:CLUSTERmirdev72
2005-06-30 10:03:01 EDT [ISS.0079.0101V2] Unable to register Reverse Connection
with alias:CLUSTERmirdev72
2005-06-30 10:03:01 EDT [ISS.0079.0100V2] Registered 0 Reverse Connections with
alias:CLUSTERmirdev72
2005-06-30 10:03:02 EDT [ART.0116.3509V2] Adapter Runtime (Notification): Notifi
cation Manager waiting on Integration Server initialization.
2005-06-30 10:03:05 EDT [ART.0116.3509V2] Adapter Runtime (Notification): Notifi
cation Manager waiting on Integration Server initialization.

Hi,

Go to the Remote Server page in the Admin console and check the Alias CLUSTERmirdev72, is the host name correct?
If you didn’t do that already, try to use the IP address for “Host” instead of a host name.
Is the port specified in the Alias available, or is it in use by another application?
Be aware: The Test button doesn’t work for aliases for Reverse Connections.

Hi wmUsers,

I have joined this group recently, i love it, coz it is very helpful for us.
so i am very thankful for the all wmUsers to provided the good resolutions…
for releted to the webMethods.

I need ur help for the below Ques…

We have configured two integrations server on different machine with Clustered.
But we are having the following exception on the start up of the Integration Server.

Error is :[POP.005.0006] This server node is not defined in the MWS cluster configuration,

thanks
Rahul

Hi wmUsers,

I have joined this group recently, i love it, coz it is very helpful for us.
so i am very thankful for the all wmUsers to provided the good resolutions…
for releted to the webMethods.

I need ur help for the below Ques…

We have configured two integrations server on different machine with Clustered.
But we are having the following exception on the start up of the Integration Server.

Error is :[POP.005.0006] This server node is not defined in the MWS cluster configuration,

thanks
Rahul