In the previous post, we discussed the why in getting User Experience insights, and where to use Stratusphere UX to fulfill this.
In this part, I am going more into the what and how part by taking Stratusphere UX for a spin. We will start with the what by doing a little architecture as a starting point.
What is needed?
The architecture of Stratusphere (FIT/UX included) has four components: Stratusphere Hub Appliance, Stratusphere Database Appliance, Stratusphere Collector Appliance, and Connector ID (CID) Key. In short, the components do the following:
- Stratusphere Hub (required), a pre-configured virtual appliance that provides the central policy management, policy distribution, data collection, reporting and alerting system for Stratusphere, as well as the user, interface for the Stratusphere user.
- Stratusphere database (required), co-installed on the Hub or separate preconfigured virtual appliance. This component provides the central data storage for the Stratusphere product line. The separated appliance is needed when having a large number of desktop deployments (500+ desktops, see the sizing guide for details).
- Statusphere Collector appliance, again co-installed on the Hub for small environments or a separated preconfigured virtual appliance that provides the ability to collect CID Key data and network monitoring data (in small environments both forms of data) for Stratusphere.
- Desktop agent, or the Connector ID Keys aka CID. Windows, MacOS and Linux Distribution in-guest agent for getting those in-guest insights about the system, applications, and user activities. The CID’s are managed by the Hub and can be installed manually, via software distribution mechanisms or integrated into the master image for virtual desktops.
Depending on the to-investigate environment the hub, database, and collector can be shared in one appliance or can be separated. Use the calculator to determine what sizing and infrastructure architecture is appropriate for your use case. The calculator can be found at http://www.liquidware.com/products/stratusphere-sizing-guide.
The following architecture diagram shows the components interacting (as taken from http://www.liquidware.com/content/pdf/documents/support/Liquidware-Stratusphere-Architecture-Overview.pdf)
Set the VDI profile for your assessment aka environment
Stratusphere will self learn and adapt the values to the assessed environment over time. However, the VDI profile is highly customizable to your needs. You can set everything according to what you (or your customer) find important for a successful UX. Start thinking about these figures when starting to define the use cases to be investigating. Speak to key users and application owners to find what is important for the user community and what values/metrics the application are dependent on. Try to find some common ground. If you put the weight of CPU higher than the weight of memory, and the applications are memory resource intensive you will have false information in your report. At that time you can work on a, or some, profile that you would wish to report on.
If this information beforehand is not clear (and this happens a lot), start with the defaults and take some more time to analyze the data.
When all is running: go to Stratusphere UX part, Diagnostics, and VDI UX Profile. You will see a screen similar to this one.
So, now we have some information about what. Let’s see about the how part and take it for a spin.
On to the Bat..erm..Demo Lab
We have a Horizon Enterprise 7.5 demo lab environment, with Instant Clone and full clone Windows 10 desktop pools. And UEM, NSX, among others. I will be deploying the Stratusphere virtual appliance here. Beforehand, use the Liquidware Statusphere sizing calculator (http://www.liquidware.com/products/stratusphere-sizing-guide) to size the appliances for the environment. This is a demo lab so I won’t require a lot.
Use the following table for the OVF deployment links:
Stratusphere Hub Appliance:
Stratusphere Database Appliance*:
Stratusphere Collector Appliance*:
After the ovf deployment checks the configuration, and start the appliance to boot up and do all the initial configurations. When the console shows, do the ALT-F2 combination.
Login with ssconsole and sspassword. From the menu, set the network configuration accordantly, set the FQDN, callback (fqdn preferable) and so on. Write configuration to finish the configuration.
And with this, the appliance is good to go. Point your browser to the address shown (or set, as I was a bit fast with the capture of above image) and login with the default ssadmin and sspassword combination and you are ready to rock. I have connected to vCenter and Active Directory for the virtual machine and user side setup.
Add clients to the environment
For Statusphere to collect some data we should add some clients and components to the environment. Next, to the valuable client information, we can also add a directory and virtual infrastructure data. Do this as well if you are investigating virtual desktop UX. Add clients by downloading and installing the CID. And let Stratusphere collect some data. For our Full Clone desktops, I have used the groups’ policy standard which I have saved to a location that can be accessed via the network (a share). Use the ADMX template that can be downloaded here also.
Create a computer policy with this template. Set the Hub Address to enabled and fill in the FQDN/IP (do this for 32 and 64-bit if you have both. And set the software settings installation package part pointing to the share.
For our instant clone desktop pools, I just installed the Windows version manually in the master image. When sealing the template run install-location\Liquidware Labs\Connector ID\admin scripts\ VMwareView_MasterImagePrep.bat. This stops the services and sets to manual. After seal create snapshot and push this image to the pool. In the post, customization run the post worker script or something else that sets the Liquidware services to auto and starts them (in order connector, UI and update). This will create a unique computer ID so we will have.
As you see in the above list, our demo environment currently contains one full clone, the base image, and three instant clone desktops.
Collect information and assess those insights
Allow some time for these products to collect and interpret the data. For Stratusphere and all similar, I advise for a minimum of a business cycle. On average a business cycle is 4 weeks or a month. The bare minimum is a few hours to two weeks. However, these results are with a little ‘it depends’. After a few hours, the first data will be there for the operators to take a look at. For example, you can use this dataset to check if the expected desktops are checking in. For assessing and analysing the data, use more time. The more measurements the more reliable the data is. Preferably plan for the period where mission or business critical processes are happening.
If you have a clear understanding, you can use the data to report on. And in a later phase, you can also integrate with systems management products or IT help desk systems via SNMP and RSS feeds.
You can find your UX profile back at for example the UX score. The UX score is a composite metric of resource consumption, and the ability to identify resource constraint limits and sort them using a weighted rating system of – Good, Fair, Poor (or A+, A, A-, B+, B-, C as labeled) depending on the UX score/profile. When your UX profile has a good match, your VDI’s will have good labels. Easy to spot.
Keep using Stratusphere UX to fully enable your organization in their digital workspace management with performance monitoring. Use Stratusphere when doing platform (re-)design, lifecycle management, and optimization. Check what impact security changes have on the UX of your users desktops, before you implement these changes. Integrate with your helpdesk system and other operational management solutions. And in many other areas.