Today I have installed the Microsoft Dynamics AX Diagnostic Framework (Beta) tool and I would like to share my first impressions with you. In simple terms it is a type of windows service, with a database and a web client. When installing the next prerequisite will pop up.
I had to struggle a little with the installation, especially to activate the correct IIS role. Log parser is downloadable from here. The easiest way is to go about this is to deploy the dashboard in the same location as where your online help is running.
When opening the browser some easy configuration steps need to be followed.
The first step is to configure a first run:
Once the first test run is complete, some handy advices appear:
The checks are a series of AX and SQL installation best practices, more closely related to AX development best practice checks. The next screen dump shows results of my installation with Contoso data. Please note that I have run this without any customizations.
When browsing through these issues, you will realise that this is default AX behaviour. It is known that there are DEL_ prefixed tables in the sys layer and that the table cache is not entirely correct on several tables.
Therefore this is a handy check for validating your installation and it also helps to pinpoint possible performance or stability issues.
Next we have some reports:
- Resource intensive queries
- Index lock contention
- Indexes most accessed
- Database IO performance
- Database wait statistics
- Missing indexes
These reports are however not new and compared to SQL Performance Dashboard reports , they are still limited. I do not understand why Microsoft have not reused these. Then we have Perfmon traces and SQL traces. I don’t know jet, if the perfmon traces can be analyzed like the PAL tool.
These are a new layer above existing tools which sounds rather promising. In fact they are a big step in the right direction. You can wave goodbye to the AXPerf tool and you no longer have to dive into the complicated world of SCOM. I also hope they will discontinue this type of logging on the IDMF tool. IDMF should stick to archiving and nothing more.
The question is do we have a tool that a customer or functional consultant will understand? The answer is NO. Let me give an example, why do we have an oil level indicator on the dashboard of our car? Simple, it is there to warn us that we should take action otherwise the car might break down. The first thing you do is to read the manual and, if you are unable to solve the problem yourself, you refer to your car mechanic. So it is easy for a car, why cannot it be just as easy for Dynamics AX?
This question has been on my mind for the past 4 years. Therefore I have created my own solution Performance Dashboard. All the information is in AX and a customer can easily monitor whether everything is fine or if he should start getting worried. The dashboard display shows the highest, lowest and average scores and also uses cues to jump straight to issues.
It also provides an overview for the day and you can also compare days. This is helpful as issues have the habit of manifesting themselves when you are not at the customer side.
This dashboard can even warn you by sending alerts or daily reports. I receive a mail every day containing an overview of the current status. These reports show me trends
It also can run best practice checks for performance issues on your tables.
The only limitation of the tool is that it cannot send an email or Alert when the AOS is down. After all though, the purpose of such a monitoring tool is to anticipate problems and avoid downtime.
More information links.