If you plan to monitor your Citrix XenApp and XenDesktop environment in order to determine availability of your infrastructure, you should use Citrix SCOM Management Pack for XenApp and XenDesktop. Monitors for checking XenApp and XenDesktop infrastructure (hosting connections, XAXD services on Delivery Controllers) and Delivery Group availability have already been available in earlier product versions. Those monitors enabled you to solve infrastructure problems quickly when they occurred and possibly even before end users would become impacted. In product version 3.9, we are adding SLA view on the infrastructure health (availability).
The following questions can be effectively addressed by using the new features in latest Citrix SCOM Management Pack for XenApp and XenDesktop:
- What is the availability of XenApp and XenDesktop infrastructure over time? At what specific times Site infrastructure was not available?
- What is the availability of a Delivery Group, and its related published applications and desktops, over time? At what specific times a Delivery Group was not available?
The latest product release introduces two new reports that enable you to get more insight into the matter. In this blog post, we will focus on features for monitoring infrastructure availability.
There are two new reports in the “Citrix Management Pack for XenApp and XenDesktop Reports“ reporting library: they are named “Site – Infrastructure Availability” and “Delivery Group – Availability”. We have already had a similar type of report: “Application – Availability”. Besides application availability, it also used to monitor Delivery Group availability which is now available in the ”Delivery Group – Availability” report.
Site – Infrastructure Availability report
This report shows availability of a Site infrastructure over time. Availability is determined by health (availability) of most important Citrix services on the Delivery controller computers (AD Identity Service, Broker Service, Configuration Service, Host Service, Machine Creation Service), and availability of configured hosting connections. For each Site you get availability graphs for each important Citrix service, for each hypervisor connection, and a graph for availability on the Site level that is calculated based on services and connections.
At the top of each report, you need to select the parameters. For each Site selected, a separate page is generated. You can also select which downtime states should be interpreted as unavailability. By default, SCOM’s own “Unplanned maintenance” is selected, which means that only unplanned maintenance will be interpreted as unavailability (besides the complete downtime of course). Because some states cannot be automatically determined whether they should be displayed as available or unavailable, you have to decide how you want to interpret them. You can also select other downtime events (Monitoring unavailable, Planned Maintenance, Monitor disabled, Unmonitored) if you wish to interpret them as unavailable. Chart scale gives you an option to check availability on hourly, daily, monthly, or yearly basis. This means that if you select, for example the “daily” chart scale, graph would show availability data grouped for each day of the reporting range. If you would later want to see more detailed availability data for a specific day, just click on the stacked column – the report scale will be set to hourly and the report duration would be set to only the selected day. The same behavior applies to other chart scales.
The report is separated into three major sections:
- Site availability
- Citrix services availability
- Hypervisor connections availability
Let’s first take a look at the main Site availability section graph. First row of data displays availability for entire reporting range in percentage and absolute time. For each (hourly, daily, monthly, yearly) entry, there is a stacked column which displays availability in color and percentage of that state. The top part of the chart is more of an overall availability state which shows available if the Site was available (in an UP state), partly available (combination of UP and DOWN states), or unavailable (in a DOWN state) in a specific time period.
The graph shows availability in percentage. Therefore, to see the exact time periods when the Site was not fully functional (was unavailable), click on “Open site downtime periods report”. This opens a new report which shows detailed downtime period information on when the Site was unavailable.
As stated previously, Site availability depends on Citrix services and hypervisor connections. A Site is considered available at a specific point in time if all services and hosting connections are available. If either of them is not available, the Site is also not available. For example, AD Identity service is not available on a Site level if it is stopped on all Delivery Controllers in the Site. When the service health is yellow (warning), it means that the service is not available in all Delivery Controllers in the Site.
The Citrix services section of the report shows availability for each important Citrix service in a Site.
As you can see, in our example only AD Identity Service was not operational and therefore Site availability is equal to the AD Identity Service availability. Other services were available for the entire time period and so were hosting connections, as you can see on the following figure.
The Connection section of the report shows availability for each configured connection in a Site.
Hypervisor connections in maintenance are treated as available. The first graph in the “Hypervisor connections availability“ section presents this state as available, but in the second graph the state is shown as yellow (warning).
Delivery Group – Availability
This report shows availability of server OS and random desktop OS Delivery Groups over the selected reporting time period. A Delivery Group is considered available at a specific time if users could start desktops/applications in that Delivery Group at that time. In other words: an available Delivery Group has at least one machine that can take a new connection (at least one machine is registered, the machine is not fully loaded and not currently in the maintenance mode).If you are familiar with »Application – Availability« report you will feel right at home with »Delivery Group – Availability« report. While »Application – Availability« report shows availability for selected applications, »Delivery Group – Availability« report shows availability for Delivery Groups. Major difference is that you select Delivery Groups instead of applications.
Delivery group availability is actually availability of desktops provided with the selected Delivery Group. It can also be used for checking application availability in case application is published by a single Delivery Groups (as in most cases). So instead of running “Application – Availability” report for each app you can run select a Delivery Group that publishes apps for which availability has to be checked. Delivery group availability is identical to application availability if application is published by single dg. In case application is published on single Delivery Group, the availability would be identical on both reports.
In situations where an application is published by multiple Delivery Groups, you should use the “Application – Availability“ report.
Delivery Group availability is tracked only for server OS Delivery Groups or random desktop OS Delivery Groups. The report should be used only for those Delivery Groups. If static desktop OS Delivery Group is selected, the report will always show it as available.
Try it out
If you are a customer with an active Platinum subscription for Citrix XenApp and XenDesktop, you can download Citrix SCOM Management Pack for XenApp and XenDesktop version 3.9 here.
If you are a user of Citrix SCOM Management Pack for XenApp and XenDesktop version 3.6 or later, you already have the data for “Delivery Group – Availability” report available. For the »Site – Infrastructure Availability« report, the data will start collecting as soon as you install the latest version of the product.