In the first article about the statistical analysis of performance data, I will be dealing with averages. I will explain why it is a bad idea to work with averages of averages (like when storing intermediate results), why relying on averages can be dangerous and what good averages look like.
As I have announced a few weeks ago, this article kicks off a new series about analyzing performance data. The individual articles will discuss different aspects of correctly working with the collected data. They will provide the necessary mathematical and statistical background for processing and visualizing the data.
In my series about custom reports for EdgeSight, you have learned about the database schema and the overall layout of a query against it as well as how the query is embedded in a report that can be displayed through the EdgeSight user interface. In all articles I have mentioned several times that reports are expressed in RDL – the report definition language which is based on the Extensible Markus Language (XML). In this posting, I’d like to take a closer look at a RDL file and provide some insight how a report is structured and what those tags mean.
Lately, I have had the pleasure to spend quite some time coding in PowerShell. I don’t think I need to tell you how much fun this is. Some things require a new approach in PowerShell but once you get the hang of it …
In the course of this coding project, I have been looking for a way to detect whether a script is running elevated. In this article I will present a script header checking for administrative rights and – if necessary – spawning a new instance of the script which asks for elevation by UAC.
In my series about custom report for Citrix EdgeSight, I introduced a lengthy example query for system performance including standard constructs for filtering by time and department as well as machine groups. Now that we know about additional parameters for process categories and user groups, let’s have a look at an example SQL query.
In part 3 of my series about custom reports for Citrix EdgeSight, I have listed some parameters that are automagically recognized by EdgeSight. Those parameters cause EdgeSight to display specialized picker or pre-populated dropdown lists. But Citrix does not provide a current documentation for report parameters processed by EdgeSight. Therefore, it’s back to reading existing reports to discover those parameters. In this article, I will introduce some more parameters for EdgeSight reports and publish a report to examing values passed from the EdgeSight web-based console to the report.
I have just released an updated version of the whitepaper about user profiles. In addition to a general overview of user profile and the challenges involved in managing them, it now includes a section about profile migration.
Performance Monitor is usually used to record performance data of one or more systems and analyze the collected data at a later point in time. As this is useful for offline performance analysis and scalability tests, real-time access to performance data is cumbersome. The graphing node of PerfMon does not offer statistical data in a useful manner. In this article, I demonstrate how to retrieve performance data for a metric and analyze it on-the-fly using PowerShell.
Microsoft has launched a new type of event called TechDay Online. It is targeted at customers with technical expertise. The first incarnation focuses on deploying Windows 7.
In today’s world, talk about virtualization and centralization is hot. Both have become proven technologies to reduce the total cost of ownership. In this article, I will show you how to achieve both virtualization and centralization while offering Windows 7 to your users at the same time using the Remote Desktop Services in Windows Server 2008 R2.
On 10.09.2010 I held a session at ice:lingen about performance monitoring and capacity management of Windows Servers. Attached you’ll find my slides.
Please note that I had to remove most of the images because I am not allowed to show them publicly.