Almost from the moment I joined Citrix Consulting several years ago, I have been part of performance and scalability testing projects both in our labs and “on the road” in consulting. Customers are always looking for accurate product performance information in order to estimate hardware requirements and costs. Partners and consultants are constantly looking for the best ways to architect a Citrix solution. And our own engineers are continually seeking to find and overcome system bottlenecks in our products.
With the acquisition of XenSource several years ago and the formation of the Citrix XenServer hypervisor, there has been a great deal of interest in the IT community in accurately monitoring XenServer for these purposes, not to mention in comparing its performance to rival products. XenServer and its management console, XenCenter, offer a refined overview of XenServer performance for every day administration. And the detailed metrics automatically gathered by each XenServer feed into advanced products that help administrators best manage their environments, such as with the Workload Balancing features of Citrix Essentials for XenServer.
But for gathering customized or more granular system performance data, customers and consultants need more: a set of knowledge not typically present in the traditional Microsoft/Citrix administrator skill set (XenServer is a GNU/Linux-based product), and a set of tools that are unavailable for niche scalability and performance testing purposes. It becomes clear that there is a need for an overview of the pre-requisite XenServer knowhow and the tools for attaining this.
This XenServer Performance Monitoring for Scalability Testing whitepaper we have put together is an accumulation of such knowledge from our recent testing efforts in the Consulting Solution Center. It provides an overview of the different ways to retrieve pinpoint performance metrics, whether they are second-by-second network data on all network interfaces on a XenServer or a running log of a virtual machine’s resource utilization during a test. Most importantly, the whitepaper provides a starting point to begin creating customized performance monitoring scripts for troubleshooting or scalability testing purposes. Topics covered in the whitepaper include:
- A sample, “quick start” XenServer performance monitoring script
- Useful commands, built-in utilities and tools for gathering performance data for the host and VMs
- Coordinating multiple XenServers during testing and retrieving the results
We hope this will better equip customers, partners and consultants in analyzing their XenServer during typical testing efforts and will help us fine-tune each of our own Porsches in the IT garage.