Tuesday, 30 December 2014

EPM - Did you know? #4

Back again with the fourth instalment of the “did you know” series and the random topic for today is:

Did you know that you can alter the frequency or disable the essbase OPMN ping?

Personally I think OPMN was a bad choice to inflict on essbase and has been more trouble that it is worth, it only seems a viable solution for clustering on *nix type systems and the rest of the time it is a waste of space, it would have been nice to be given the option of not having to deploy it with essbase, anyway it is an old technology and looks like it is not going to be in Oracle’s strategic future.

In case you are wondering what the OPMN ping is all about I will give a little bit of background information.

If you look in the essbase product bin directory you will see the OPMN ping executable:

What it basically does is log into essbase with the admin account and then log out this indicates whether essbase is alive or not which is a requirement when clustering.

By default this happens every 20 seconds and if you take a look at the essbase log you will see it constantly polluted with something like:

[Tue Dec 30 10:12:29 2014]Local/ESSBASE0///964/Info(1042059)
Connected from []

[Tue Dec 30 10:12:49 2014]Local/ESSBASE0///964/Info(1042059)
Connected from []

As you can imagine this amounts up to many entries over time and doesn’t serve much purpose in the log.

There are also other logs that get written to under the OPMN logs directory.

I seem to find that at many clients this log folder is not part of any housekeeping solution and the essbase ping and agent log can become large in size.

Unfortunately there does not seem to be any options to rotate these logs even though there is a setting in the opmn.xml

The rotation size works with the opmn log but not does not seem to work with the ping or agent log, I remember a enhancement request back in 2012 to be able to add the option to but I am not sure it got anywhere.


If you look in the EssbasePing.log then you will see the following messages repeating every 20 seconds.

[2014-12-30T11:02:15.272-18:02] Initializing Ping Request
PM-PingUtility INFO: Connection is done
PM-PingUtility INFO: Received Response from the Essbase Agent
PM-PingUtility INFO: Essbase Message - PING_OK

If there is a problem connecting or logging into Essbase then you will see a similar message to:   

[2014-12-30T11:25:13.539-18:58] Initializing Ping Request
PM-PingUtility: Couldn't connect to Essbase with Essbase ERR# 1042006
PM-PingUtility: Fails to Ping the Essbase Agent 1042006

In a non-clustered essbase environment this is not really providing much benefit as the message just repeats and unless the log is being monitored it would go unnoticed, if required the same solution could be easily created with a Maxl script.

Luckily the opmn ping log does not have a lock on it so it can be removed with stopping services.

So how about altering the frequency of the ping and how can it be done.

Well it is back to the opmn.xml file located in:


There is an element called ping and an attribute called interval that can be added to the XML file

For example <ping interval="60"/>

The interval is the number of seconds between each ping so in the above example after a restart to OPMN the ping will have gone from 20 seconds to 60 seconds.

After making any changes to the xml file I was always like to use the validate command to check the file is still valid.

The maximum value is 7200 seconds and if you try and set it any higher then you should see the following error message:

The maximum value is defined by the XML schema definition file:

It is possible to change the value in the XSD file so I tested increasing the maximum value to 10800 and the ping interval matched the value.

So how about disabling the ping process well it couldn’t be simpler just set the interval value to 0.

If you set it to 0 and restart OPMN then you will no longer be bothered by the ping.

Saturday, 27 December 2014

EPM - Did you know? #3

Today I am going to be continuing the “did you know” series with the third instalment.

This time the randomiser has picked:

Did you know where the EAS console picks the version/patch number from?

I previously wrote a blog on understanding product versions in workspace, in the post I did not cover EAS and I know sometimes there can be confusion on the versioning in the console so here is a brief overview on how it works.

In the console if you go to “Help” and select “About Essbase Administration Services” then the following dialog box will open.

As you can see it displays three product version numbers, I have no idea why Business Rules is still included as it was removed in and this causes the version to default to 4.1.1.

The first question is it the server version for EAS and APS and the answer is no it is only the version of the files being used in the console.

You can tell that by the MSI console version as you don’t even need to connect to the EAS server for it return the version and even when you connect it does not make any difference.

With the web console it downloads the latest versions of the java files from the EAS server so usually the console and server will be in sync.

So where is the version generated from, well both the web and MSI console use a java archive called eas_client.jar, the MSI version is located in


The web console downloads the file from the EAS server and the file is contained in the EAS web application file:


If you open the java archive there is a directory which contains property files and the AboutDialog.properties has some of the information we are after.

The VersionNo property contains the information that it is displayed in the console.

To prove it I updated the version and then loaded up the console again.

So that covers the EAS version so how about the APS version.

This is generated from a different java archive ess_es_gui.jar which is also used by the console:

The version is derived from the manifest file in the META-INF directory, the manifest is a special file that can contain information about the files packaged in an archive file and in this case the version is read from the Implementation-Version line.

Once again I updated the version number.

Loaded up the console again and the version had changed.

You may sometimes find that with patch releases that the jar files are not updated so the version number does not match to the patched server version.

Well at least you should now understand how the versioning works in the console.

Wednesday, 24 December 2014

EPM - Did you know? #2

If you read yesterday you will know I am putting together a series of quick posts with EPM information which you may or may not know.

The randomiser today has picked:

Did you know you can simply automate the refreshing of the Shared Services cache using the Java API?

In the Shared Services under the user directory configuration there is the option to set the cache refresh interval which the default is 60 minutes.

In some cases with large corporate directories the interval may be set to higher value to reduce performance degradation.

You may be wondering what the cache is for anyway, well here is Oracle’s definition:

"Interval (in minutes) for refreshing the Shared Services cache of groups to users relationship data. Default is 60 minutes.

Shared Services caches information about new external user directory groups and new users added to existing groups only after the next cache refresh. Users provisioned through a newly created external user directory group do not get their provisioned roles until the cache is refreshed."

Now say there have been changes in the external directory such as users have been added or removed from groups then these will not be picked up until the next cache refresh, if these changes are required to be pushed through straight away or maybe changes should be applied at a set time then having to log into Shared Services to click the refresh button is not the ideal option.

Luckily there is an option available using the API and a simple piece of code will allow the refresh to occur on demand.

Here is an example piece of code I wrote to basically accept a username and password which is then authenticated and if successful the cache is refreshed.

Apologies to master coders out there if you are appalled by my attempt :)

Right, let’s demonstrate refreshing the cache with an example scenario, in Shared Services I have an active directory group called EPM_ADMINISTRATORS which has already been provisioned with roles.

At present the group only has only user called epm_admin, in the AD I add myself to the group

Checking the properties of the group in Shared Services still only shows one member.

This will be the case until the next refresh of the cache so time to give my code a quick test.

Running the java can easily be done from command line with the following:
  • Set the classpath to include the required java classes, to simplify I am using epm_j2se.jar which references many packages
  • Set the epm instance location
  • Call the java class passing in the epm instance and user credentials.
So for my environment I used:

set classpath=.;E:\Oracle\Middleware\EPMSystem11R1\common\jlib\\epm_j2se.jar;

set EPM_ORACLE_INSTANCE=E:\Oracle\Middleware\user_projects\epmsystem1

java -DEPM_ORACLE_INSTANCE=%EPM_ORACLE_INSTANCE% refreshCache  admin password

The warnings can be ignored as they not important, I would recommend adding some encryption though as clear text credentials are not for the security conscious.

Checking the properties of the group in Shared Services confirms the refresh has happened with myself now displaying as a member.

If you are using essbase and run the above code the caching system essbase uses will not automatically trigger an update until a user logs in, so you could update the code to include a login using the essbase JAPI or Maxl and then you should see entries in the SharedServices_Security_Client.log to confirm the cache refresh did happen.

Well that completes #2, until next time..

Tuesday, 23 December 2014

EPM - Did you know? #1

I thought I would try something different over the festive period and as it is a time of sharing I am going to put together a series of very quick posts on EPM related information which you may not know..

I have created a list of topics and for each blog I will randomly pick one from the list, now they may be of no interest to you or you may find you already know them but I am sure some will find them useful, I am not sure how many I will get through as it is all time dependent but I am aiming for five.

Today the randomiser has chosen:

Did you know there is an undocumented tool that allows to change the data location for essbase databases?

The tool is command line driven and can change the location without the need to export/import or restructure the database.

It can be used to quickly change the data location on any database, it can be used in migrations where the drive structure is different between source and target and if done correctly can be part of an upgrade process from 9.3 onwards.

If you have ever worked with the essbase staging tool then you will know there is an archive file available in the EPM installation structure.

The tool can be found in the Migration folder and is used as part of the upgrade process for essbase.

If you extract the archive and look in the bin directory there is an executable file called essmove (available on both Windows/*nix)

As I said there is no documentation available but I found out about it by accident and then had a play around with it, it is called as part of the staging tool but can be used in isolation.

To be able to use the tool you first must set a number of environment variables but don’t worry there is a script already available in the Essbase server bin directory, look for the setEssbaseEnv script.

Once the script has been run then execute essmove and follow the onscreen questions though make sure the intended database has been stopped first.

In the above example the drive location for the ind/pag files has been moved from E to F.

The changes are applied to the databases kernel file (*.esm)

If you check in EAS or with Maxl you will see the location has been updated.

Before starting up the database make sure the index and data files have been copied to the new location.

The tool can be used against ASO databases in the same way.

In the above example the default tablespace location has been updated to F:\Essdata\App

The changes are applied to the metadata tablespace file.

If you check in EAS or by Maxl then you should see the location has been changed.

I have used the tool on many occasions and have found it invaluable, it is also possible to pass values into the tool so it be part of a scripted process.

So that completes the first did you know, I hope you found it useful.