• View Communities
    • Citrix Developer Network
      The place for unfiltered straight talk on Citrix products. Blogs, code downloads, best practices, APIs, and more can all be found here.
    • Citrix Ready Community Verified
      Does it work with Citrix? Application compatibility questions are a thing of the past with the new Citrix Community Verified site.
    • Blogs
      Learn the latest from the Citrix employees who are building application delivery infrastructure technologies.
    • Blogosphere
      The Citrix Blogosphere is a window into the thousands of conversations taking place about Citrix and Application Delivery.
  •  Sign In
The Citrix Blog
Blogs for tag 'scalability'

Permalink | Twitter Post to Twitter | Comments (0) | Views (726) |

posted by Kimihiko Kitase

Recently I was asked about the performance and scalability of XenApp from our CSA. And I got the following reports or informaitons. It will be helpful to everyone to check these reports and informations at single site.

XenApp 5.0
XenApp 5 Scalability Analysis
http://support.citrix.com/article/CTX119108

Citrix Presentation Server 4.5
Scalability - Number of Users per Presentation Server
http://support.citrix.com/article/CTX114848

Effects of Varying the Number of CPUs of a Citrix Presentation Server
http://support.citrix.com/article/ctx114844

Citrix Presentation Server 4.0
Citrix Presentation Server for Windows Server 2003 x64 Editions - Performance and Scaling Capability
http://support.citrix.com/article/CTX107341

MetaFrame XP
How Hyper-Threading Affects User Capacity of Metaframe XP Servers
http://support.citrix.com/article/ctx101880

Advanced Concepts Guide for MetaFrame XP with Feature Release 3 - P79
Effects of Varying the Number of CPUs and Hyper-Threading on MetaFrame XP Servers
http://support.citrix.com/article/ctx102497

MetaFrame 1.8
Windows 2000 Terminal Service Capacity and Scaling
http://support.citrix.com/article/ctx9354

Other sources
http://h71019.www7.hp.com/ERC/downloads/4AA0-9214ENW.pdf
http://www.dell.com/downloads/global/power/ps1q07-20070244-Citrix.pdf
http://www.brianmadden.com/blogs/brianmadden/archive/2004/02/19/the-4gb-windows-memory-limit-what-does-it-really-mean.aspx

Test tools
Citrix Server Test Kit Users Guide 2.1
http://support.citrix.com/article/ctx101744

EdgeSight for Load Testing 3.0
http://support.citrix.com/product/es/ltv3.0

Expand Blog Post
Permalink | Twitter Post to Twitter | Comments (0) | Views (7074) |

posted by Florian Becker

Dan Feller on my team contributed at least two posts on the topic of virtualizing XenApp servers on XenServer. Dan makes some excellent points and gives you plenty of business reasons why XA on XS is a good idea.

I am not going to re-iterate Dan's points here, but rather focus on another burning question in this context: How much of a scalability overhead can I really expect with my specific application? The typical consulting answer would be "it depends" and "we'll have to do a scalability / performance assessment to determine the specifics and best practices". So, we have done just that and used two popular enterprise class Applications: Siebel 8.0 and PeopleSoft 9.0. The Solution Center is one of the teams under the umbrella of Worldwide Consulting Solutions (Dan Feller's Integrated Solutions team is another) and focuses on these types of projects, which often involve third party applications and/or hardware platforms from our technology partners.
Recently, we looked at running the front-end of Oracle's PeopleSoft and Siebel applications on XenApp (both 32-bit and 64-bit platforms) and focused on comparing the user densities we could achieve on "bare metal" servers compared to running them on XenServer.
The results are published in two separate whitepapers (PeopleSoft, Siebel), which describe the test bed, test methodology, detailed results and interpretation. As Dan stated in his May 15th posting, the virtualization overhead can be as low as 6% for XenApp virtualization on XenServer, and our tests confirm this number. Of course, the numbers vary between the applications and platforms, and we describe all the details in the whitepapers.
Generally speaking, kernel memory limitations constitute the first bottleneck on 32-bit platforms, and our tests verified that behavior. Even with the popular /PAE switch, the kernel memory limitation remains at 2 GB. Therefore, you can expect a higher user density per physical server if you're running multiple 32-bit XenApp servers on a XenServer. You'd have to be cautious not to consume too many CPU cycles, which often become the next bottleneck once memory is no longer a major concern. Prices of multi-core, multi socket servers with plenty of RAM have come down significantly, so chances are that your latest servers have plenty of resources to run reliably in that configuration at a reasonable price:

According to this 1988 article, prices of 1 MB memory chips were as high as $60 (or $105 in today's money), while you can buy a barebones server with 64 GB of RAM for roughly $5,000 today. While I am on the topic of computer nostalgia: a 150 MB hard drive set you back over $8k in today's dollars way back when... 1988 was also the year Dan Feller was looking forward to seeing his favorite TV show getting its own slot in the line up and he is still enjoying it to this day, as you can see from the quotes in his postings on this site. But I am digressing...

The Solution Center also conducted detailed validation tests with Oracle to obtain validation status for running virtual images of the Web-, Application-, and Database servers of Siebel 8.0 , PeopleSoft 9.0, and Oracle E-Business Suite 12 on XenServer 4.1, so you can now be confident that the entire environment can be successfully virtualized on XenServer, allowing you to take advantage of XenMotion in case of hardware failure and other benefits.

Expand Blog Post
Permalink | Twitter Post to Twitter | Comments (1) | Views (11738) |

posted by Daniel Feller

The last time I wrote about XenServer and XenApp, I focused on a whole set of items like manageability, availability, flexibility and utilization.  This time, I want to focus directly on utilization as based on the feedback I've received it seems it's the one many people are interested in.

Even before the scalability numbers of XenApp and XenServer came out, I had numerous conversations about virtualizing XenApp.  And now that Citrix is showing the XenServer overhead for virtualizing XenApp, those conversations have increased, but I think some critical points are being lost.  A couple of months ago, Citrix did scalability tests to identify that XenServer has roughly a 7-8% overhead when virtualizing 64-bit XenApp, and roughly 20% when virtualizing 32bit XenApp servers.  I was like WOW, 64bit is great, barely any overhead.  But how many people are actually running a truly 64bit environment?

Most people have the hardware, as it has been sold for years. Most people also have access to the 64bit version of Windows and XenApp. So why aren't we all jumping on the 64bit bandwagon? Because it's the applications. Unfortunately, many applications that XenApp environments run are 32bit, and some are still 16bit!  This conversion to 64bit applications will take time (Does anyone else remember the 16bit versus 32bit migration that happened years ago? It wasn't an overnight thing. It took time. And yet there are still 16bit apps out there.)  So this fact makes it highly unlikely that organizations will be able to convert their XenApp environments into complete 64bit setup. This means many will stay with 32bit only or else have mixed 32/64 bit environments.  So let's focus on the 32bit environments, are they virtualization candidates?

Maybe And most likely Yes.

Take a look at many XenApp deployments and what resource do you typically exhaust first?  RAM.  It is because in Windows 2003, we are limited to 4GB of addressable RAM.  So, when we hit that limit, everything else in the system is wasted (processor, IO and networking).  And I've seen some applications take enormous amounts of RAM. Just the other day I was working on some detailed Visio drawings and Visio took 1GB of RAM.  Yes, I said gigabyte.  (Of course the drawing was about the Simpsons and how Homer stays at the forefront of technology - He even had a blog called "Mr X. - All the Muck That's Fit To Rake").  So, a 20% overhead on 32bit systems? I probably wouldn't notice as my entire server is barely utilized except the RAM.  

RAM is easy to install and one of the cheapest things to add to a server.  Use the same hardware and increase the RAM to at least 8GB.  Now, try to run 2 virtual XenApp servers.  You might not double your user concurrency, but you will get pretty close, which will equate to hardware and power savings.

So take a look at your physical XenApp servers. Is the RAM fully utilized? What about the processor utilization levels?  I bet more likely than not the RAM is fully committed and the processors are running at 10-40% utilization. 

Daniel

(Homer Simpson Quote: "I want to share something with you: The three little sentences that will get you through life. Number 1: Cover for me. Number 2: Oh, good idea, Boss! Number 3: It was like that when I got here.")

Expand Blog Post