Xen VS OpenVZ Round Two Continued - Updated 7-2-07 By Trevor Connors published: Wednesday, May 09 2007
Pradeep Padala (author of HP Labs "Performance Evaluation of Virtualization Technologies for Server Consolidation") responds to questions about the differences between Xen and OpenVZ.
On april 25th VSM published a paper found on HP labs that was simular to one we published earlier in the year regarding the comparison between Xen and OpenVZ. This article found here was later updated with comments from Kir Kolyshkin from SWsoft as well as Simon Crosby of XenSource. I felt that the author of the HP paper should also be included in any questions arsing from his work so I contacted him to seek his input. Below is my corrospondance with Pradeep.
Trevor: Pradeep, Thank you for taking the time to reply to us at VSM.
After reading the paper found on the HP labs website I was not clear on a few things so I have a few more questions for you if you do not mind.
First I must admit you paper was very interesting and do not think anyone will refute your outcome but I am wondering if you believe that these two technologies do the same thing and should be back to back in the same evaluation?
Pradeep: I am not sure what you mean by back to back evaluation. As mentioned by a few people in the comments to your article above, the two technologies have their place. Hypervisor-based technologies have a different performance-isolation tradeoff than container technologies such as OpenVZ, and which technology one prefers depends on what you want.
In our context, we are comparing two virtualization technologies to understand how they can be used for stacking multiple multi-tier applications in consolidation scenarios. Our goal is not to conclude which one is better (which may not even be possible). We are thus evaluating the overheads posed by the different virtualization technologies in order to understand how to "right-size" the containers within the context of virtualization. Our results demonstrate (not unexpectedly) that hypervisor-based technologies have a higher overhead, most likely as a result of L2 cache misses. If the different applications do not require complete isolation, then clearly something like openVZ would be preferable from a performance point of view. However, if isolation is paramount, or if the applications being stacked require different OS configurations/OS's, then one has to use a hypervisor based technology. In both instances, however, the consolidation exercise requires a quantification of the virtualization overhead from an application's perspective (e.g., response time or work done) , which is the subject of our paper.
The paper also points to an area of future research -- namely how can hypervisor-based technologies be streamlined to make better use of L2 caches and reduce the performance hit.
Trevor: Can we have an evaluation of these products "simultaneously running RHEL 3, Windows Server 2003 and Vista"?
Pradeep: I guess you are referring to evaluating Xen in above configurations. It would be an interesting study, but our focus is mainly on understanding the fundamental differences between os-level and hypervisor-level virtualization technologies. As we found out, the big difference is the amount of L2 cache misses. We would like to further investigate what design/implementation choices cause this difference.
Trevor: Finally, do you believe that the capability of hosting different OS's is not important and evaluating this aspect would not be worthwhile.
Pradeep: That's an interesting question. It very much depends on the requirements of customers/users. Some customers may need to run multiple OSs in the same node, while others might just want to host multiple applications running on the same operating system. Xen is useful in the first case and OpenVZ is better for the latter.
Trevor: Thank you once again Pradeep, I look forward to your upcoming article and future research on the streamlining of hypervisor-based technologies "to make better use of L2 caches and reduce the performance hit."
Reply from Simon added on 5-18-07
Xen can host multiple Operating Systems and OpenVZ cannot. OpenVZ can host bazillions of tiny VMs that share the same core OS code and Xen arguably can't do that efficiently. I could tell you, without any research, that if 90% of the code is the same you'll have better cache performance.
Right from the start, I've said what Xen is good for, and what it isn't good for. Given that we all agree, can we just move on?
Updated on 7-2-07
Today I found this post on Pradeep's Blog site. http://ppadala.blogspot.com
"Our tech report received 10,000 hits in just the month of May. See a screenshot of stats here. Yay!
Our recent tech report on performance evaluation of Xen vs. OpenVZ is generating quite a buzz. It's the second most viewed tech report from outside HP. Virtual Strategy Magazine picked up on it and wrote an article. A few people including Simon Cosby, CTO of Xensource responded. Slashdot picked it up yesterday. I am pleasantly surprised that there is not much FUD generated by the Slashdot readers in the comments. Some of the comments are quite informative."
"Let me point out a few facts straight away. Both technologies have their place. Xen trades off performance for security and isolation and if you want to run multiple operating systems Xen or VMware is the only way. OS level container technologies like OpenVZ beat these hypervisor-based technologies easily on performance, but they may not be appropriate for some situations."
The slashdot listing is here below are some samples
Virtual Strategy Magazine has also recorded new Podcasts with both XenSource and SWsoft
Listen to Xen Summit Update Here
Listen to OS Virtualization Vs Hypervisor Virtualization Here
Both are good.(Score:5, Informative)
by Blackknight (25168) on Monday May 14, @10:30PM (#19125023)
(http://www.watters.ws/)
As somebody that actually has experience setting up and running virtualized systems (I work for a web hosting company) let me add my two cents here.
OpenVZ is ok if all of your child environments run the same OS and you don't care about them stealing each other's resources. We constantly have problems with customers overloading their VPS and causing problems with the other environments, this doesn't happen with Xen. I've fork bombed child environments and caused the load to spike to over 700 until it crashed, dom0 and the rest of my domUs just kept running like nothing was even happening.
OpenVZ also wins if you want to oversell hardware, Xen doesn't have "burstable" memory like OpenVZ does. Personally I prefer Xen for the jailing that it does and you can also run multiple OSes at the same time. I have a server at work that's running CentOS, Windows 2003, Windows Longhorn, and Gentoo all at the same time, OpenVZ only lets you run Linux on Linux.
Xen vs VMware - personal experience(Score:5, Informative)
by Anonymous Coward on Monday May 14, @11:41PM (#19125517)
Here are a few talking points based on my experiences with both Xen and VMWare (ESX Server & Workstation).
1. Cost - no contest, xen wins hands down $0 vs $5000/cpu.
2. Performance - xen wins noticably, i can get away with running 8 virtual machines with 1/4 the hardware that VMWare required for 6.
3. Capabilities - VMware versions ESX GSX and up beat Xen in ease of use and flexibility, anything less and Xen wins. Xen does have a quicker live migration capability, but falls short on conencting external hardware to the virtual machine (something that is trivial in VMware).
4. Stability - about even, maybe Xen. I've seen ESX crash once, and have never seen Xen crash.
5. Ease of use - VMWare no question. Theres a learning curve to Xen and setting up new VM images is a pain. (which is disappearing) There are fantastic tools for VMware that will let you manage virtual machine creation and even migration from a physical box. (some of which can be used to feed xen too =))
The bottom line. A Xen setup (using open source version) that can easily run 12 VMs costs about $2,000 on DIY hardware. A VMWare server/software combo to do the same? about $30,000. Assume an additional 40-80 hours of learning to get up to speed with Xen vs VMware.
The comments above cover the open source / free version of Xen. I have no experience with the commercial Xen offerings.
Personally i'm Xen biased due to cost.
Re:Both are good.(Score:5, Informative)
by flydpnkrtn (114575) [pfloyd AT nixwizard DOT net] on Monday May 14, @11:27PM (#19125415)
(http://nixwizard.net/)
When you sit down and do your homework, and don't oversell as you said, OpenVZ gets the job done. I was looking for a good way of virtualizing a new server we were moving to and I've gotta say I can't see myself ever NOT installing a server in either an OpenVZ VE or a Xen domU ever again.... with OpenVZ you can give one VE (almost) all the resources of the hardware its running on, and when that one server outgrows that hardware doing a vzdump or vzmigrate is sooo easy.
The big difference between Xen and OpenVZ comes down to what openvz calls "privvmpages" - memory that is claimed by running processes but not actually used.
|
