[history] VMM Quadrants

[history] VMM Quadrants

[Rich Persaud]

I am new to VMM internals, but have experience in database (Sybase) 
internals and data center infrastructure.   RDBMS engines that run on 
many OSes go out of their way to reimplement OS facilities (e.g. memory 
management, threads, disk and network I/O) to insulate their performance 
from OS intermediation.  

The dispelling of newbie ignorance can sometimes yield good diagrams for 
future newbies.  Do you have recommendations for research papers with 
diagrams/history of virtualization implementations?  I'm looking for 
high-level overviews with conceptual models that are low-level 
accurate.  Two overview papers:

Denali, VMware, User-Mode Linux (is Fig. 3 correct? where would Xen go?)

   http://www.eecs.umich.edu/techreports/cse/2002/CSE-TR-465-02.pdf

VM/370, Denali, Xen:

   http://www.robertwrose.com/vita/rose-virtualization.pdf

Is there a definitive bibliography of web-accessible virtualization papers?

Rich




-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click

Re: [history] Virtualization Papers

[Rich Persaud]

[-- Attachment #1: Type: text/plain, Size: 341 bytes --]

Following up on my earlier message, I've attached an HTML index of 
papers on virtualization (best viewed with a CSS-capable parser).  

Started this to understand the historical context of Xen, but it may 
speed the orientation of new developers.

Please advise if I missed anyone or if any "Topics" could be more 
precisely named.  

Rich

[-- Attachment #2: xen_virtualization_research.html --]
[-- Type: text/html, Size: 9705 bytes --]

Re: [history] VMM Quadrants

[Martin Maney]

On Fri, Mar 04, 2005 at 03:01:01PM -0500, Rich Persaud wrote:
> Denali, VMware, User-Mode Linux (is Fig. 3 correct? where would Xen go?)
> 
>   http://www.eecs.umich.edu/techreports/cse/2002/CSE-TR-465-02.pdf

I don't remember seeing any answer to this (which may say more about my
memory than what actually happened), but since I was browsing through
the spool, looking for seed material for wiki pages, I was just looking
at this.  I may be wrong, but I'd guess that Xen 1.2 would go about
halfway between VM/370 and Denali (like both it directly and solely
controlled the hardware, but it supported more of the real architecture
for its guests than Denali, eg., virtual memory).  Given only two
dimensions, I suppose Xen 2 has drifted somewhat to the left, as it has
offloaded most of the hardware management onto dom0.  Perhaps it would
be more accurate to add a third dimension and place X2.0 out of the
plane of the graph (to represent the fact that the hardware-managing
host is actually one of its virtual machines - a rather level-confusing
approach) as well as about as far right as VMware.  Xen-unstable is
confusing things even more, probably.

(I'd forgotten about your followup with the longer list of papers.  Is
that available on the web somewhere that I could link to?)

-- 
Allen Funt was one of the great psychologists of the twentieth century.
His informal experiments and demonstrations on "Candid Camera"
showed us as much about human psychology and its surprising limitations
as the work of any academic psychologist.  -- Daniel Dennett



-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click