Service Center end to end monitoring
Proof of Concept
Two vSphere hosts are monitored with all virtual machine on them, the License server, Zone Data Collector, Netscaler and Storefront so that we have a full view of the Citrix environment. We added a Catalyst and a firewall to the list and started the monitor on monday, two days ago. Within a few minutes data was flowing in and after two days we get a nice view of the environment and it’s issues.
Because of the application processes reporting issues, the layer above it shows a color so I checked if they were related. The Windows Service layer is reporting a lot of error messages so let’s open that one.
As you can see it fairly easy to dive into the server and see what is happening there. No need to connect to the server to look at the eventlogs, eG can handle that from one console just fine.
The next server to look at was the data collector was it was also showing a red color. it says that there is an issue with server availability. Let’s take a closer look.
I had to wipe out all the server names for their customer won’t like it if they are mentioned here I guess. What we noticed was that six servers were listed there that are not present anymore in the Citrix farm. So somehow the data collector is getting info from somewhere that these server should be there. This is currently being investigated for that is not what you want. To be clear, the six servers in red are orphan Citrix servers of a long lost past 😉
As we looked that the graph it shows this VM does this every day so perhaps it’s worth looking at what it’s doing.
So I checked if the overall CPU performance of the ESX host was impacted by this. I looked at the graph for the CPU performance.
The overall CPU performance of the host was not too bad but comes close to 90% often, perhaps some reorganizing is needed here. Again under investigation with the customer, they have some homework to do.
Looking at the graph for this user it seems like a glitch as it went up and down fast. These graphs are very useful for service desks for they give instant information about the users session.
Another thing it monitors is the profile size of the user, the customer still uses Roaming profiles (what??? yes they know and are going to change). The user has a profile size of 200MB and it is growing as Roaming profiles grow. There are more users there with this large profiles so they know they need a UEM solution to fix this… work in progress.
We only monitored two days but already we can see the growth in the profile for this user, going up by the hour. that’s why you need a decent UEM solution.
Also on the application side of things you can see the number of instances being used on that server and the CPU usage of the application. This customer had issues with Winword.exe being aggressive on CPU and they wanted to know who was using it at that time. With all the tools they had now they had no way getting that data, now they can.
A report asked from many customers within Citrix environment is about logon performance. With eG there is a report for that, it shows the logon duration for the users. these reports are customizable so that they will show what you are looking for. I chose the default TOP7 one to show that it monitors logon times. Citrix has a nice break down of this itselfs which I like a lot.