Monday, January 24, 2011

How to configure authentication in Quests QMX for Websphere WAS

If you try to configure Quests Management Extension for Websphere Application Server 6/7 you realize very soon, that there is no possibility to provide username or password for the Performance Monitoring Servlet URL.

With the following workaround you are able to connect to the servlet but unfortunately you have to enter the password in clear text. So at least it may be a good idea to change the file permission to a minimum requirement for the files affected. And it would be good to use a dedicated service user with the “monitor” permission in WAS.

There are two files (called PMIServletReader.js and TestConnection.js) to be modified in the folder %ProgramFiles%\eXc Software\WMI Providers\nonWindows\Virtual Agent Library\MOM\Websphere

The important script is PMIServletReader.js where you have to modify line 99. There you find the following string:

l_objXMLHTTP.open("GET", strURL_in, false);

Add the following (bold) text to the line and replace username and password to the one you like to use:

l_objXMLHTTP.open("GET", strURL_in, false, "username", "password");

Double-check that the quotation mark (") is not a special character. I prefer to delete the copied ones and replace them by the one given by the editor.

Save the edited script. If you like to test the connection (what I recommend) you have to edit the same line (maybe other line number) in the second script called TestConnection.js. Now you are able to test the connection within Quests Configuration Utility by right click on the WAS server entry and selecting Test Connection. Verify that you have enabled the server by unmark the "disabled” checkbox before testing. Don’t forget to save the change.

You should also be sure, that the value in the field “Websphere Application Server Name” in Quests Configuration Utility is exactly the same as in the XML given by the servlet. Simply open the URL for the servlet in your web browser and copy the value from the tag called “Server name” and paste it into the field from above. This must not be the name of the server itself.

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

Friday, January 21, 2011

Updated my Microsoft Certifications

Now I successfully updated my MCSE, MCSA and MCPs to the newest MCITP and MCTS (for SCOM too)!

Read Full Post...

Wednesday, January 19, 2011

Deploy SCOM agent to untrusted zones with Opalis

Normally I publish my own thoughts on my blog. But because the Opalis workflow from Anders is very interesting and a good example how Opalis can help you to automate administrative tasks I will post the link to his blog: http://contoso.se/blog/?p=2054

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

Tuesday, January 18, 2011

Wrong percentage values in Quests QMX for VMware?

After installing the Management Extension from Quest for VMware vSphere (get it for free from here) you’re getting alerts from percentage counters? Further, you figured out that the values are above 100%?

First of all: are you sure that you installed the Framework using ENU regional settings? Please take note that only this language is supported from Quest. This could be a possible problem because of wrong decimal symbol (comma instead of dot).

If the language settings are correct, you can verify the data in the XML files written from the QMX module. Therefore find a file called *_HostSystem.xml in the following path and verify the language in the tag mDesc:

%ProgramFiles%\eXc Software\WMI Providers\nonWindows\Virtual Agent Library\MOM\VMware\vCenters-vSpheres\Log

In case that this language is not English you have to change it. Please take note that the following steps must not be supported. If you need a supported way contact your VMware support.

  1. Rename the folder you’ll find in step 2 (in my case de for German but can be any other language) to *_old:
  2. %ProgramFiles%\VMware\Infrastructure\VirtualCenter Server\locale\de
  3. Make a copy of the en folder and rename it to the language as founded in step 2.
  4. Restart the virtual center server.

After that you can start the module for vSphere again (if not set to start agent automatically) and verify the new configuration in tfhe XML as described above.

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

Tuesday, January 11, 2011

How to install Quest QMX for Websphere MQ

Honestly, the documentation of Quests Management Extensions could be much more better. So as for the Extension for Websphere MQ Series.

Here you find the main steps how to install the QMX, in this case in a clustered environment running on Windows Server 2008 R2.

  1. Install SupportPac on every MQ Server Node. Therefore run the following setup from the server with QMX base frame framework and MQ extension installed: %programfiles%\eXc Software\WMI Providers\nonWindows\Virtual Agent Library\MOM\MQSeries\IBMEventsMonitorSupportPacForWindows\setup.exe
  2. Install Telnet Server Feature on every MQ Server Node.
  3. Create a domain user. Add the user to the local group “TelnetClients” and to the local group “mqm” on all nodes.
  4. On the active node start a CMD and run the following command: runmqsc
  5. In the CLI run the commands (see step 6, 7 and 8) from the Quest Configuration Tool provided at the configuration site for MQ for every (!) queue and channel you have configured. You need the and even if they are not shown in the IBM documentation
  6. ALTER QMGR MONQ(MEDIUM) MONCHL(MEDIUM)
  7. ALTER QL('Q1') MONQ(QMGR)
  8. ALTER CHL('QM1.TO.QM2') CHLTYPE(SDR) MONCHL(QMGR)
  9. In the Quest Configuration Tool encrypt the username and password for the domain user account from step 3. Enter the username in the format domain\username
  10. Now add the filepaths to the username and password field at the configuration site for MQ
  11. Change the server name to the correct one. In a clustered environment use the cluster alias and verify, that you can reach the cluster with that. Otherwise take the IP (as I had to do because of customer configuration)
  12. Change the port to telnet (23)
  13. Safe your configuration
  14. Test the connection by right-click on the server name and select Test connection
  15. Now configure the channels, events and queues as you need it for your business
  16. Uncheck the Disabled checkbox and safe the configuration
  17. Start the monitoring by right-click on the server name and selecting Start
  18. Start the tracing console by right-click on the server name and select Show log
  19. There you can see if the extension works as expected.

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

Friday, January 7, 2011

How to remove an Action Server from Opalis

Normally, when you like to remove an Action Server from Opalis you use the Deployment Manager. Simply right-click the server to be decommissioned and select Uninstall Action Server.

But sometimes it happens, that the Action Server is not available any more when you try to do that. The step above ends with an error message (“Failed to uninstall the Action Server” with detail “RPC not available”).

In that case there is no supported way to remove the Action Server from the Management Server. But you can remove the Action Server by using the following three SQL statements:

SELECT [UniqueID]
  FROM [Opalis].[dbo].[ACTIONSERVERS]
  WHERE [Computer] = '<action_server_to_be_removed>'
GO

DELETE
  FROM [Opalis].[dbo].[ACTIONSERVERS]
  WHERE [Computer] = '<UniqueID_from_1st_query>'
GO

DELETE
  FROM [Opalis].[dbo].[OBJECTS]
  where UniqueID = '<UniqueID_from_1st_query>'
GO


Replace the string <action_server_to_be_removed> by the server name exactly as provided in the Deployment Manager. Further, replace the <UniqueID_from_1st_query> by the UniqueID you get from the first statement.


All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

Monday, January 3, 2011

Error 0x80004005: Unspecified Error

You deployed an Action Server through the Deployment Manager without any issue but it doesn’t appear in the GUI?

And you are getting this error and the following in the logs provided in C:\Program Files (x86)\Opalis Software\Opalis Integration Server\Action Server\Logs (or without x86 on 32bit OS) on the new Action Server:

First log file:

<Param>This computer was unable to communicate with the computer providing the server.</Param>
<Param></Param>
<Param>-2147467235</Param>

Second log file:

<Param>Unspecified error</Param>
<Param>[DBNETLIB][ConnectionOpen (Connect()).]SQL Server does not exist or access denied.</Param>
<Param>-2147467259</Param>

You worked the way described in the manual how to deploy an Action Server automatically?

And further you checked the steps in the knowledge base?

In my case I deployed the Action Sever through the Deployment Manager but I could not start the Action Server service. After I configured the DB as in the manual described in the section “Manual Deployment of an Action Server” it works as expected and the new server appears in the Deployment Manager under Action Servers.

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

See policy names of workflows to be triggered

When you create a new nested workflow with more than one policy included you may find out, that when you want to select the next policy in your workflow that one will always be named “New Policy X” even when you renamed the policy to what you like it (the X stands for none or any other number).

To get the correct names you just have to check in (and out again if you want) the policy once.

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...

The trick how to export new workflows…

Maybe you’d like to export your newly created workflows to have a backup in case that you are modifying them to death later?

Did you verify, that the workflow objects are contained? You should!

However, to get the objects into the export you must (!) check in the policy first. Otherwise you will have nothing but the folder within your export.

Bad, but that’s the fact…

All information is provided "as is" without any warranty! Try in lab before. Handle with care in production.

Read Full Post...