Bug 781013 (SOA-3493)

Summary: Internal materialized views are not refreshed on both nodes in a Single, Non-Multihomed, Server
Product: [JBoss] JBoss Enterprise SOA Platform 5 Reporter: Paul Nittel <pnittel>
Component: EDSAssignee: Van Halbert <vhalbert>
Status: CLOSED NOTABUG QA Contact:
Severity: high Docs Contact:
Priority: high    
Version: 5.2.0.ER5   
Target Milestone: ---   
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
URL: http://jira.jboss.org/jira/browse/SOA-3493
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Fedora 14
Last Closed: 2011-11-15 17:48:47 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:

Description Paul Nittel 2011-10-17 15:57:08 UTC
Steps to Reproduce: Install EDS in production profile
Copy profile to node1 and node2
Copy datasources and VDB to both deploy directories
Start both servers
Query cluster: SELECT * FROM PartsVirtual.SupplierInfo
Query each node: SELECT * FROM SYSADMIN.MatViews
Note the LoadState column and Updated timestamp
Wait a minute
Query cluster: SELECT * FROM PartsVirtual.SupplierInfo
Query each node: SELECT * FROM SYSADMIN.MatViews
Note the LoadState column and Updated timestamp
On each node CALL SYSADMIN.refreshMatView('PartsVirtual.SupplierInfo',false)
Query each node: SELECT * FROM SYSADMIN.MatViews
Note the LoadState column and Updated timestamp
project_key: SOA

I have a cluster which is configured as a two-node single, non-multihomed, server. On each node (deployed as "Production" and named "node1" and "node2") I have deployed identical datasources and VDB.

Shouldn't Materialized View management take place on all nodes in a cluster?

The servers are started using:
nohup ./run.sh -c node1 -g EDSPartition -b IPaddress -u 239.255.100.100 -Djboss.Domain=EDS -Djboss.jvmRoute="node1" -Djboss.messaging.ServerPeerID=1 -Djboss.service.binding.set=ports-default &>node1.out &

nohup ./run.sh -c node2 -g EDSPartition -b IPaddress -u 239.255.100.100 -Djboss.Domain=EDS -Djboss.jvmRoute="node2" -Djboss.messaging.ServerPeerID=2 -Djboss.service.binding.set=ports-01 &>node2.out &

I connect to the VDB using the following URL:
mm://IPaddress:31000,IPaddress:31100

Querying the IMV table (PartsVirtual.SupplierInfo) loads one of the nodes, but not both. I can refresh both nodes by CALLing SYSADMIN.refreshMatView on each node in the cluster.

Additionally, the transformation being queried contains a TTL of 30 seconds which never appears to expire. (It works as expected in a non-clustered environment.)

Comment 1 Van Halbert 2011-10-17 15:57:08 UTC
Link: Added: This issue Cloned from TEIID-1782


Comment 2 Van Halbert 2011-10-17 15:58:01 UTC
Security: Added: Public


Comment 3 Van Halbert 2011-10-18 18:58:34 UTC
This turned out to be a user error.