SocialMiner Agent Request - Connection to CCE issues

From DocWiki

Revision as of 14:33, 12 December 2013 by Linichol (Talk | contribs)
Jump to: navigation, search

Connection to CCE issues

Problem Summary MR PIM won't go ACTIVE. SocialMiner's Connection to CCE Notifications server status displays something other than "Established connection with CCE Media Routing PG" when you hover over it.
Error Message

1 The Connection to CCE Notifications server status indicator displays "CCE Configuration for Multichannel Routing is disabled" when you hover over it.

2 The Connection to CCE Notifications server status indicator displays "Listening for incoming connection from CCE Media Routing PG" when you hover over it. The SocialMiner logs don't reveal anything, but the MR PIM log may help.

If the SocialMiner IP address is actually 10.86.141.169 and the port configured on SocialMiner is the default 38001.

For an incorrectly defined SocialMiner Hostname/IP Address in MR PG Setup, the MR PIM log may look something like this:

 14:58:11:428 PG2A-pim1 Trace: Transport: Invalid hostname or IP address: 10.86.136.999
 14:58:11:428 PG2A-pim1 Invalid MR application IP address / host name "10.86.136.999".

-OR-

 15:04:04:693 PG2A-pim1 Attempting to connect to MR application at IP address 10.86.136.111 port 38001.
 15:04:14:584 PG2A-pim1 ProcessPIMSetIdleReq: Peripheral 5005 going idle.

For an incorrectly defined port in MR PG Setup, the MR PIM log may look something like this:

 15:19:57:037 PG2A-pim1 Attempting to connect to MR application at IP address 10.86.136.169 port 38004.
 15:20:07:428 PG2A-pim1 ProcessPIMSetIdleReq: Peripheral 5005 going idle.

3 The Connection to CCE Notifications Server Status indicator displays "Listening for incoming connection from CCE Media Routing PG" when you hover over it.

If the MR PG hostname is actually "bos-w2k8-pg2" and port 38001 is defined in Peripheral Gateway Setup.

For an incorrectly configured MR PG Side A host (misconfigured as "foo"), the SocialMiner runtime log may show something like this:

 0000000478: 10.86.136.169: Oct 03 2013 15:50:24.095 -0400: %CCBU___________MRSS-3-MR_SERVER_SOCKET_MANAGER_ERROR: %[exception=java.net.UnknownHostException: foo]: Exception encountered while validating a host against white list MRServerSocketMgr_1380825947241
 0000000479: 10.86.136.169: Oct 03 2013 15:50:24.095 -0400: %CCBU___________MRSS-3-EXCEPTION_INFO: %[build_date=Dec 31, 1969 7:00 PM][build_type=rel][exception=java.net.UnknownHostException: foo


For an incorrectly configured port (38004 on the SocialMiner side, 38001 on the MR PG side), the SocialMiner runtime log may show something like this:

 0000000545: 10.86.136.169: Oct 03 2013 15:56:04.662 -0400: %CCBU___________MRSS-3-MR_SERVER_SOCKET_MANAGER_ERROR: %[exception=java.net.SocketException: Socket closed]: Server Socket Connection closed.
 0000000546: 10.86.136.169: Oct 03 2013 15:56:04.662 -0400: %CCBU___________MRSS-3-EXCEPTION_INFO: %[build_date=Dec 31, 1969 7:00 PM][build_type=rel][exception=java.net.SocketException: Socket closed

Invalid (in-use) port configured:

 0000002226: 10.86.141.220: Oct 04 2013 14:36:29.252 -0400: %CCBU___________MRSS-6-MR_SERVER_SOCKET_MANAGER_ENTER: %[MR_THREAD_NAME=MRServerSocketMgr_1380911789252]: MR ServerSocketManager starting. Started a thread
 0000002227: 10.86.141.220: Oct 04 2013 14:36:29.256 -0400: %CCBU___________MRSS-3-MR_SERVER_SOCKET_MANAGER_ERROR: %[exception=java.lang.NullPointerException]: Unexpected exception occurred. Continue to process other requests
 0000002228: 10.86.141.220: Oct 04 2013 14:36:29.260 -0400: %CCBU___________MRSS-3-EXCEPTION_INFO: %[build_date=Dec 31, 1969 7:00 PM][build_type=rel][exception=java.lang.NullPointerException
 at com.cisco.ccbu.ccp.mr.ServerSocketManager.process(ServerSocketManager.java:392)
 at com.cisco.ccbu.ccp.mr.ServerSocketManager.call(ServerSocketManager.java:95)
 at com.cisco.ccbu.ccp.mr.ServerSocketManager.call(ServerSocketManager.java:47)
 at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303)
 at java.util.concurrent.FutureTask.run(FutureTask.java:138)
 at com.cisco.ccbu.infra.threads.InstrumentedRunnable.run(InstrumentedRunnable.java:92)
 at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886)
 at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908)
 at java.lang.Thread.run(Thread.java:619)
 at com.cisco.ccbu.infra.threads.ThreadPoolThread.run(ThreadPoolThread.java:164)
 ][product_name=CCP][subsystem_exception_info=][tid=MRServerSocketMgr_1380911789252][version_number=CCP_9_0_2_0_0_0_803]:   Information associated with the following logged exception


4 Both CCE and SocialMiner are configured properly, but CCE still does not connect to SocialMiner or continually connects and disconnects.

Possible Causes

1 CCE Configuration for Multichannel Routing is not enabled on the System Administration gadget of the Administration tab in SocialMiner.

2 MR PG (side A and/or side B) is set up improperly in Peripheral Gateway Setup.

3 CCE Configuration for Multichannel Routing is misconfigured on the System Administration gadget of the Administration tab in SocialMiner.

4 Potential issues include high network latency and/or low network reliability. This could be caused by misconfigured DNS.

Recommended Actions

1 Enable CCE Configuration for Multichannel Routing and verify and configure the rest of the Media Routing PG Configuration.

2 Fix the Hostname or IP Address or Port in MR PG Setup.

3 Fix the Hostname or IP Address or Port in the CCE Configuration for Multichannel Routing in SocialMiner.

4 See SocialMiner Administration - potential delay during DNS lookup SocialMiner runtime logs including "MR_SERVER_SOCKET_DNS_WARN" indicate that it has taken longer than 5 seconds to validate the host, and will include the total time to have done so. SocialMiner runtime logs including "MR_SERVER_SOCKET_DNS_HOSTNAME" and "MR_SERVER_SOCKET_DNS_CANONICAL" document how long it took to look-up an individual hostname or FQDN, respectively.

Discussion: When a host attempts to connect to SocialMiner, the IP address of the incoming connection is compared to the list of allowed servers. DNS and reverse DNS lookup are used to attempt to match the incoming connection to a hostname in the the A/B host configuration. If DNS is misconfigured and SocialMiner is not able to lookup the hostnames for the incoming connection and the IP addresses of the specified A/B hosts, the connection will be rejected. If DNS is misconfigured in such a way that the lookup is possible but takes an excessive amount of time, the connection may be allowed, but the MR protocol negotiation may timeout, causing the connecon to be closed.

Also note: if there are multiple MR PGs set up to point to SocialMiner (for example, two different Side A MR PGs) and both are concurrently attempting to connect to SocialMiner, they will continually “fight” each other and neither of the connections will persist.

Release 10.0(1)
Associated CDETS # None

Rating: 5.0/5 (1 vote cast)

Personal tools