|
Problem(Abstract) |
Collecting data for problems with the IBM® WebSphere®
Application Server versions 5.0 , 5.1 ,6.0 and 6.1 for synchronization
component. Gathering this MustGather information before calling IBM
support will help you understand the problem and save time analyzing the
data. |
|
|
|
Resolving the
problem |
 |
If you have already contacted support, continue to the component-specific
MustGather information. Otherwise, click: MustGather:
Read first for all WebSphere Application Server products.
Note: Collecting the following MustGather information has been
automated in the IBM
Support Assistant product feature for WebSphere Application Server.
For more information about this automated data collection feature, see Using
IBM Support Assistant to collect MustGather data.
Synchronization specific MustGather information
This technote explains the information needed to better understand
why specific files are not properly transferred and synchronized or why
the administrative console or wsadmin display a node as being Not
Synchronized.
Do the following to test and collect the correct information for
understanding synchronization problems:
- If you are running versions 5.0 or 5.1, make sure the File
Transfer enterprise application is running on the deployment manager
(dmgr) by checking the running applications in the console.
- If it is installed but stopped, start it.
- If it is not installed, reinstall
the File Transfer application.
- Try to do a full resynchronization of the problem node. This double
checks if synchronization is initialized. The following are two ways to
perform a full resynchronization:
If the node status is Unknown or the MBean is not returned, this
is not a synchronization problem, but instead a discovery problem.
- Using wsadmin commands:
- Identify the ConfigRepository MBean and assign it to variable
a:
wsadmin>set a
[$AdminControl completeObjectName type=
ConfigRepository,process=nodeagent,node=nodename,*] |
|
=>Returns a MBean. |
|
- Invoke the object operation without any parameter:
wsadmin>$AdminControl invoke $a
refreshRepositoryEpoch |
|
=>Returns an integer. |
|
- Set the variable b for node synchronize:
wsadmin>set b
[$AdminControl completeObjectName type=
NodeSync,node=nodename,*] |
|
=>Returns a MBean. |
|
- Synchronize by issuing the following command:
wsadmin>$AdminControl invoke $b sync |
|
|
where: |
is: |
set |
JACL command |
a,b |
Variable names |
$ |
JACL operator for substituting a variable name with its
value |
AdminControl |
Object that enables the manipulation of MBeans running in
a WebSphere server process |
completeObjectName |
$AdminControl command that creates a string representation
of a complete ObjectName value based on a fragment |
process |
Key property that represents the name of the server
process in which the object runs |
node |
Key property that represents the name of the node on which
the object runs |
type |
Key property that indicates the type of object that is
accessible through the MBean |
invoke |
Invokes the object operation without any parameter.
Returns the result of the invocation. |
refreshRepositoryEpoch
|
Resets the node digest set |
sync |
Attribute of modify objects |
|
|
- Using the administrative console:
- Expand the System Administration section and select
Nodes.
- Check the box next to the nodes that are not synchronized.
- Click Full Resynchronize.
- Trace the synchronization process.
- Enable synchronization tracing on the dmgr:
- Edit the server.xml file for the dmgr located in the
following directory:
$WAS_ND_HOME/config/cells/cellname/nodes/nodename/servers/dmgr |
|
- Find the TraceService XML tag section.
- Modify the startupTraceSpecification entry to look like the
following:
- For version 5.0 and 5.1:
- For version 6.0:
- Make sure the enable entry is set to true.
- Recycle the dmgr server and clear the ffdc and dmgr log directories.
- Use the syncNode script on the Base install:
- Stop the nodeagent. This is done because the syncNode script
uses the nodeagent SOAP port.
- Execute the syncNode script located in your
$WAS_BASE_HOME/bin directory.
- Command with security Disabled:
syncNode.bat(sh)
cell_hostname dmgr_SOAP_port -trace |
|
- Command with security Enabled:
syncNode.bat(sh)
cell_hostname dmgr_SOAP_port -username
uid -password
pwd -trace |
|
- Run the collector tool on both ND and BASE.
- If Step 3 works and you still have synchronization problems with your
nodeagent, trace the nodeagent synchronization process with the dmgr.
- Enable synchronization tracing on the dmgr:
- Edit the server.xml file for the dmgr located in the
following directory:
$WAS_ND_HOME/config/cells/cellname/nodes/nodename/servers/dmgr |
|
- Find the TraceService XML tag section.
- Modify the startupTraceSpecification entry to look like the
following:
- For version 5.0 and 5.1:
Sync=all=enabled:Admin=all=enabled |
|
- For version 6.0:
- Make sure the enable entry is set to true.
- Recycle the dmgr server and clear the ffdc and dmgr log directories.
- Enable synchronization tracing on the nodeagent:
- Edit the server.xml file for the nodeagent located in the
following directory:
$WAS_BASE_HOME/config/cells/cellname/nodes/nodename/servers/nodeagent |
|
- Find the TraceService XML tag section.
- Modify the startupTraceSpecification entry to look like the
following:
- For version 5.0 and 5.1:
Sync=all=enabled:Admin=all=enabled |
|
- For version 6.0:
- Make sure the enable entry is set to true.
- Recycle the nodeagent and clear the ffdc and nodeagent log
directories.
- Run the collector tool on both ND and BASE.
What to do Next?
Once you have collected the preceding information, you can begin Analyzing
the data or simply submit
the diagnostic information to IBM support.
Other published IBM documents to review on this issue:
For a listing of all technotes, downloads, and educational materials
specific to synchronization, search the WebSphere
Application Server support site. |
|
|
|
|
|