In this guide I walk quickly through the installation and a sample configuration for collecting windows data, performance data and creating some simple reports.
Taken straight from the beta readme file here is the list of changes from the previous version:
•Added Server and Cloud Enrollment scenario
•Added Remote Desktop scenario
•Added Capability to inventory software using SWID tags
•Added option to open reports folder from report generation progress window
•Updated SQL Usage Tracking scenario to include information collected through UAL
•Updated System Center Configuration Manager Usage Tracking scenario to include user data collection
•Updated Windows Azure VM Readiness assessment to include all supported versions of Linux
•Updated Windows Azure VM Capacity assessment to include assessment for A5 Azure VMs
•Removed third party browsers from Internet Explorer assessment. Third party browsers are still visible in the installed products view if detected
•Removed LAMP stack from Linux Environment scenario
•Removed MySQL and Sybase migration feature scenarios
Click I Accept the terms in the License Agreement. Then click Next
Choose your install location then click Next
Select whether to Join the customer experience program or not then click Next
Then Click Install
MAP 9.0 Beta is now installed.
Launch the Microsoft Application Planning Toolkit
Give the database a name e.g. “demo customer 1”
Enter a description “Server Consolidation”
Then click Perform an inventory
Now select the required scenario’s (you can select many)
I’m just going to select Windows Computers and Windows Volume Licensing
As I’m running this on a standalone machine I’m going to manually enter computer names and credentials
Next Click Create and enter credentials
As you can see I’ve used a local account format
We now have the option to specify credential order, for this demo I’ll leave this.
Now we need to enter a computer name and account, click Create
In this scenario I have configured to use the all computer credential
Type the computer name and click Add
Now click Next
We now have the option to modify PowerShell connection properties, for this example (I know I haven’t got a CA) I’m going to disable SSL validation.
We are now given a summary of our configuration, if happy we can click Finish
We can now see our data collection task
From a process explorer view we can see MapToolkit.exe and sqlservr.exe processes
I’ve expanded the details on the data collection window so we can see a detailed view on the progress.
Our single machine collection has now completed so we can click Close
I’m now going to show that client detection has shown that I am running a desktop
One odd thing to note is that I’m running Office 2013 so I would expect that to be shown.
Since we are demonstrating server virtualization we now go to the server virtualization menu
As we can see here we have no performance data. So now we click Collect Performance Data
Since this is a demo I’m going to let this only run for 30 minutes.
Select the targets
We know that we are going to use the all computer credentials, here we can see that performance data will be collected WMI. Click Next
Verify the credential order and click next
If happy with the configuration click Finish
We can now see the task status
We can close this if we want to continue using the tool. The task engine with MAP can only schedule and run a single task at a time.
I’ve had a look and there is still no service created by the MAP toolkit, this exposes an issue that the user session running MAP and the map toolkit.exe process must remain active throughout the collection period. Considering the length of time an assessment is usually run for (31+days), avoiding interruption to this is key to keep the monitoring task as hands off as possible.
I can demonstrate this by trying to exit MAP when the task is running:
Clearly during data collection we want to avoid this situation.
I recommend that the following controls are put in place:
- Run MAP on a dedicated, newly created, clean, fully updated virtual machine.
- Restrict access to the MAP virtual machine
- Where possible try to use locate on a clustered environment
- Monitor the host and guest and configure an alert if the server is restarted
- Remotely (if possible) check in on the progress on a regular basis
Now we will wait for data to be collected.
There are a number of reports that map can generate. These come in the format of Excel Workbooks.
To generate reports is slightly more in-depth than older versions of MAP. Based on the combination of analysis configuration and where you attempt to run a report will provide a different select of report options.
Note: You can’t generate reports while the task processor is busy, so if you have to conduct a long analysis then I would suggest performing some preliminary inventory reports prior to conducting performance data collection.
I’m not going to list every report as that would generate pages and pages of screenshots. I will however demonstrate how to get the inventory data report.
Navigate to the Environment Menu
Click on Inventory Results
Now Click Generate Inventory Results Report
We can now see a folder with a single excel document in it.
Opening this reveals the following report:
As we can see this is hardware based report.
Moving back to the Server virtualization menu we see the following:
We now are ready to run the Server Consolidation Wizard
We now have the option to start modelling our target hosts.
I’ve shown the sample host configuration.
Now we need to look at consolidation ceilings. 100% would be very aggressive and leave no room for growth or spikes so normally we should limit this. However for this example I’ve taken the defaults and clicked next.
Again we will choose from a list. Click Next
I’ve selected my sample machine then click next
Review the summary then click finish
Now we click can see there is a server consolidation scenario with 1 host and 1 machine chosen. Click on this
Now we click on Server Virtualization Report
We now have a report which we can review:
As we can see to virtualize my single machine we require only one host matching our sample host profile.
Next we can look at running a private cloud fast track Wizard
Click Run Private Cloud Fast Track Wizard
We now have the option to select our target environment
Modify as required then click Next
Again we have limiting options, again I will take the default values, click next
Again I will select my single computer so click next
Review the summary then click finish
We can now see we have 1 machine and 1 private cloud in the private cloud fast track scenario, click on this scenario
We now have the option to generate a report
As we can see creating a private cloud in this example is not required however this should provide enough guidance to get you familiar with the tools use and its outputs.
In the older version of MAP you could click generate reports and create them en-mass which found was very useful, however in later versions you have to specify the exact report area to generate a report. It’s a good idea to generate all the reports possible.
I’ve used MAP for a number of years and found it a very useful tool. It does however have its limitations these to my mind mainly sit in the following areas:
- The tool lacks a service therefore is impacted by process termination, logoff events and reboots. This can limit the tools effectiveness in larger environments.
- It’s agentless, which in some ways is great, in others it leaves gaps for data collection. You can’t for example use Map effectively for VDI analysis.
- It seems the word reports have been removed. While I can see why, these were often useful for proposal data, you can however manually create what is required from the excel documents.
- Its scalability seems limited, sure you can deploy multiple instances but that makes it cumbers on to manage.
- Application data is consolidated, if as part of a discovery exercise you want to know what app is on which machine, out of the box this is not available. I’ve always intended to investigate the SQL database but to date have not had sufficient reason to go to this level. I’ve utilized other reporting mechanisms to retrieve this data (WMI scripts, vbs and management tools such as SCCM)
As a point solution MAP fits, for quick fire discovery, SMB and presales use its great. For other uses I would look to leverage other toolsets.