Tweaking Linux and WINE for World of Warcraft

It’s great that the World of Warcraft runs well on recent editions of WINE(Unbuntu/Fedora), but even when using OpenGL the typical Linux user can expect a frame rate reduction from 15%-30% over Windows XP.  However, I must point out that there are many reports of frame rate drops in Windows Vista and Presumably Windows 7(because of virtualization of XP apps).

All this is okay for typical grinding, leveling, and five man instances.  That said,  my GeForce 7950 GTX is only cranking out 30 FPS in low populated zones…  So for the bigger raids, PVP, and highly populated zones like Dalaran it can be frustrating trying to pwn at 12-16 FPS.  In fact my mission is to get at least 40 fps with the graphics card I have.

I was able to improve my frame rate by disabling mods like ‘Questhelper’ in battle grounds.  I added additional RAM to my system.  Although the frame rate boost was negligible; loading times were cut in half. (more…)


Fedora 11 will likely not include Xen Dom0 (virtualization) support

xen logo

I (like many of you) have been patiently waiting since Fedora 8 for Dom0(Domain 0) Xen support in Fedora.  Why hasn’t Red Hat or the Fedora Project made an announcement? Haven’t we been good? I mean libvirtd is great and all, but Xen PV(paravirtualized) VM’s destroy. I did some googling to get to the bottom of this. I found a fedora project page with a January 2009 status update.

Here is a snipit:

“Currently, the Fedora kernel-xen package is based on forward-porting of the XenSource patches from 2.6.18 to more recent kernel versions. This has many problems, including:

  • XenSource code has no chance of being merged upstream, in the near future, making the forward-porting work needed for all new kernel versions.
  • Lots of porting work for each new kernel version
  • Because of the above, kernel-xen has been some releases behind the non-xen kernel package, and the lag between kernel and kernel-xen has been increasing constantly”

And also:

“As of November 2007, the kernel-xen forward-porting was being finished for 2.6.22, and Linux 2.6.24 was about to be released. The effort needed for 2.6.23, 2.6.24 and later would have been even bigger with the introduction of paravirt_ops and the i386-x86_64 merge upstream. Thus, the decision was made to abandon the forward-porting effort and focus on upstream paravirt_ops.”

So where does this leave us?  Unmodified guest are old news.  Even Microsoft can do that.  Well not really, as I understand it, Microsoft’s HyperV platform contains Xensource licensed code.  But a customer of the company I work for likes HyperV a lot(incidently).  On the Xen Wiki it says that Paravirt_ops will be ported to the 2.6.30 kernel.  My prediction, Xen Dom0 support will be available toward the end of Fedora 11’s cycle or Fedora 12.

Why does Dom0 matter?  Dom0 is the specially modifed Xen-linux kernel that sits on top of the hypervisor. From Dom0 you can run fully virtualized guest and partially virtualized guest (paravirtualization).  Paravirtualized guest enjoy a method for allowing the use of a set of generic virtual device drivers provided by Dom0.  PV guest are known to have outstanding perfomance compared to their fully virtualized counterpart.  Paravirt_ops refers to Dom0 integration with the Linux kernel.


Cisco to enter the Virtualization Server Market

Is this Cisco's future?

I just read on that Cisco has announced a server line of products.  I find this interesting because the main reason I don’t use Cisco products is their high cost.  They make some of the best switches and routers, but do they have the mentality to be competitive in the mainstream server market?
One of the huge selling points of virtualization technology is the ability to cut power and hardware cost.  Knowing Cisco their baseline server product will probably be blade server costing $10,000 or more(+ whatever VMware’s cut is).

here is a quote from the CNN article:

“Chambers says virtualization is one of his big priorities for 2009, along with globalization, video growth, customer relationships, and Web 2.0. With its strong position among corporate information technology departments – Cisco has 61% of the router market – the company surely will be able to get an audience for its servers. But if he aims to beat HP, Dell, and IBM on their home court, Chambers is going to have to serve up something truly groundbreaking.”

You can read the whole article here.

The New York Times also ran an article that brought up some excellent points.

“The product — a server computer equipped with sophisticated virtualization software — is a bold but risky move by Cisco into an unfamiliar, intensely competitive market that typically produces far lower profits than Cisco makes from network gear”

You can read the whole article here.

I seriously doubt Cisco is ready to compete against HP, IBM, Microsoft and the Open Source Community.  Virtualization has taken years to become a reality.  Cisco will rely on a VMware product that buyers can already find on cheaper x86 hardware.  By contrast the VoIP market is ripe for the picking, escpicially if Cisco prices more aggressivly against open source Asterisk.  Gianormous Corporations never learn…