Skip to main content

VMware does a performance study on AMD's RVI

Nice read, this doc.

In a native system the operating system maintains a mapping of logical page numbers (LPNs) to physical page numbers (PPNs) in page table structures. When a logical address is accessed, the hardware walks these page tables to determine the corresponding physical address. For faster memory access the x86 hardware caches the most recently used LPN->PPN mappings in its translation lookaside buffer (TLB).

In a virtualized system the guest operating system maintains page tables just like in a native system, but the VMM maintains an additional mapping of PPNs to machine page numbers (MPNs). In shadow paging the VMM maintains PPN->MPN mappings in its internal data structures and stores LPN->MPN mappings in shadow page tables that are exposed to the hardware. The most recently used LPN->MPN translations are cached in the hardware TLB. The VMM keeps these shadow page tables synchronized to the guest page tables. This synchronization introduces virtualization overhead when the guest updates its page tables.
Using RVI, the guest operating system continues to maintain LPN->PPN mappings in the guest page tables, but the VMM maintains PPN->MPN mappings in a second level of page tables, called nested page tables. When a logical address is accessed, the hardware walks the guest page tables as in the case of native execution, but for every PPN accessed during the guest page table walk, the hardware also walks the nested page tables to determine the corresponding MPN. This composite translation eliminates the need to maintain shadow page tables and synchronize them with the guest page tables. However the extra operation also increases the cost of a page walk, thereby impacting the performance of applications that stress the TLB. This cost can be
reduced by using large pages, thus reducing the stress on the TLB for applications with good spatial locality.

For optimal performance the ESX VMM and VMkernel aggressively try to use large pages for their own memory when RVI is used.


Get the Whitepaper here
Labels:

Comments

Post a Comment

Popular posts from this blog

Security: VMware Workstation 6 vulnerability

vulnerable software: VMware Workstation 6.0 for Windows, possible some other VMware products as well type of vulnerability: DoS, potential privilege escalation I found a vulnerability in VMware Workstation 6.0 which allows an unprivileged user in the host OS to crash the system and potentially run arbitrary code with kernel privileges. The issue is in the vmstor-60 driver, which is supposed to mount VMware images within the host OS. When sending the IOCTL code FsSetVoleInformation with subcode FsSetFileInformation with a large buffer and underreporting its size to at max 1024 bytes, it will underrun and potentially execute arbitrary code. Security focus
Post a Comment
Read more

Splunk that!

Saw this advert on Slashdot and went on to look for it and found the tour pretty neat to look at. Check out the demo too! So why would I need it? WHY NOT? I'd say. As an organization grows , new services, new data comes by, new logs start accumulating on the servers and it becomes increasingly difficult to look at all those logs, leave alone that you'd have time to read them and who cares about analysis as the time to look for those log files already makes your day, isn't it? Well a solution like this is a cool option to have your sysadmins/operators look at ONE PLACE and thus you don't have your administrators lurking around in your physical servers and *accidentally* messing up things there. Go ahead and give it a shot by downloading it and testing it. I'll give it a shot myself! Ok so I went ahead and installed it. Do this... [root@tarrydev Software]# ./splunk-Server-1.0.1-linux-installer.bin to install and this (if you screw up) [root@tarrydev Software]# /op...
Post a Comment
Read more