Muli Ben-Yehuda's Journal
[Most Recent Entries]
[Calendar View]
[Friends]
Below are 20 journal entries, after skipping by the 20 most recent ones recorded in
Muli Ben-Yehuda's LiveJournal:
[ << Previous 20 -- Next 20 >> ]
| Thursday, June 26th, 2008 | | 5:28 pm |
recent happenings and MMCS 2008 position statement
I'm in lovely Boston this week. Originally I was supposed to attend the Xen Summit North America 2008, MMCS 2008, and USENIX 2008, but I ended up missing the Xen Summit completely and making a just-in-time appearance at MMCS, due to some trouble at home. Thankfully everything is fine now. At the Xen Summit my colleague Todd DeShane presented our preliminary results of a quantitative comparison of Xen and KVM. I'm sorry I missed the summit, by all accounts I heard it was a pretty fun event. At USENIX, I presented a poster on our SCIMITAR work which introduces the notion of "virtual memory for I/O devices". It got generally positive feedback, including a few "wow, that's pretty cool" comments. I kept wishing our prototype was at a stage where we could finally have some results, but as I told those who asked, if we had had any results, I would've been presenting a paper, not a poster... At MMCS, I participated in a panel on "Platform Management---Coordination or Not?", moderated by Intel's Milan Milenkovic. Platform management is a subject that comes up fairly often in my day-to-day work, so it was fun spending some time thinking about it and trying to come up with an appropriate position statement for the panel. The way I see it, the primary function of a panelist is to encourage the audience to think, and the easiest way to do that is to be a bit controversial. A panel where everyone---both panelists and audience---agree on everything, is a dull affair. Here is my MMCS 2008 panel position statement---paraphrased, not transcribed! For context, you probably want to read the panel description first. "Hello everyone, Thank you for inviting me to participate in the panel. When thinking about the questions Milan raised with regards to platform management, I had a bit of a problem coming up with a suitable position statement. As it happens, I either wear or used to wear many different hats. I used to work on real applications. I currently work on operating systems and hypervisors, and occasionally I dabble in computer architecture. I wondered, what kind of insight could I get from each of these perspectives? From the application writer's perspective, there are three different classes of application writers: those who do not care about the platform at all, those who think they know everything best but actually don't, since they only have localized knowledge, and those that really do know everything best. From the operating systems developer perspective, you don't get too many kernel developers who don't care about the platform, so we can leave those out. That leaves two classes: those who think they know best, but are wrong since again they only have localized knowledge (think virtual machines), and those who really do know best. From the hypervisor's developer perspective, well, really, hypervisors are just a new name for operating systems! From the computer architecture folks perspective, well, they just know everything best. Having said that, they have a whole lot of legacy stuff to support which severely constrains what they can and want to do. Now, having shared with you all of these perspectives, let me share something else: none of it actually helped me come up with a good answer to the question's Milan raised. So I tried a different approach. I assume most of you are familiar with the end-to-end argument in system design, which is usually applied to networking. Well, what does the end-to-end argument tell us about platform management? As far as I can apply it, it tells us that we should let each layer do what it knows how to do best. Not a very satisfying answer. Well, let's try a different approach then. Programming language folks like to say that programming languages should make "the simple things easy, and the hard things possible". And here I think we may have actually hit on something. How does this apply to platform management? The way I interpret it, it means two things. First, that each layer in the system should make sensible decisions on its own. Second---more importantly, and hopefully somewhat controversially---that each layer should provide a "chicken switch". What's a chicken switch? It's a term you sometimes hear in hardware design. It means that when you have all these new functionality in your design, which is not as well tested (or thought out...) as the old, reliable stuff, you usually enable it (it's new and shiny after all), but you also provide a way of "shouting chicken". A way of disabling the new stuff and going back to the old, proven way of doing things. By analogy, I claim that the most important feature a platform component can have when it comes to management, is a way for the layer above or below it to tell it to "do nothing, and get out of the way". Thanks you." | | Monday, May 26th, 2008 | | 1:30 am |
| | 1:17 am |
| | 12:44 am |
| | Sunday, April 27th, 2008 | | 11:00 pm |
I spent a couple of pleasant hours today trying to wrap my head around KVM's MMU code. After reading "The Shadowy Depths of the KVM MMU", it suddenly starts to - almost - makes sense. The tentative agenda for the 2008 KVM forum has been published and is chock-full of awesomeness. | | Sunday, March 23rd, 2008 | | 12:03 am |
The Charge of the Light Brigade, by Alfred Tennyson
Half a league, half a league,
Half a league onward,
All in the valley of Death
Rode the six hundred:
'Forward, the Light Brigade
Charge for the guns' he said
Into the valley of Death
Rode the six hundred.
'Forward, the Light Brigade!'
Was there a man dismay'd?
Not tho' the soldiers knew
Some one had blunder'd:
Theirs not to make reply,
Theirs not to reason why,
Theirs but to do and die,
Into the valley of Death
Rode the six hundred.
Cannon to right of them,
Cannon to left of them,
Cannon in front of them
Volley'd and thunder'd;
Storm'd at with shot and shell,
Boldly they rode and well,
Into the jaws of Death,
Into the mouth of Hell
Rode the six hundred.
Flash'd all their sabres bare,
Flash'd as they turned in air,
Sabring the gunners there,
Charging an army while
All the world wonder'd:
Plunged in the battery-smoke
Right thro' the line they broke;
Cossack and Russian
Reel'd from the sabre-stroke
Shatter'd and sunder'd.
Then they rode back, but not
Not the six hundred.
Cannon to right of them,
Cannon to left of them,
Cannon behind them
Volley'd and thunder'd;
Storm'd at with shot and shell,
While horse and hero fell,
They that had fought so well
Came thro' the jaws of Death
Back from the mouth of Hell,
All that was left of them
Left of six hundred.
When can their glory fade?
O the wild charge they made!
All the world wonder'd.
Honour the charge they made!
Honour the Light Brigade,
Noble six hundred!
| | Wednesday, March 5th, 2008 | | 10:45 am |
| | Monday, March 3rd, 2008 | | 10:22 am |
notes from the ASPLOS 2008 poster session
Here's a short list of posters from the ASPLOS '08 combined opening reception and poster session that I found particularly interesting. "MemCrawler: Discovering Structures in Memory" - Given a dump of physical memory, a list of interesting patterns and a list of object constraints, MemCrawler will identify kernel data structures and functions in the dump. "Coevolution of Operating Systems and Asymmetric Single-ISA CMPs" - Not all cores and pieces of code are created equal - is it feasible to run specific pieces of operating system code (e.g., the TCP/IP stack) on smaller and less complex cores? Our ASPLOS paper explores a similar question. "Multi-host I/O sharing by using I/O virtualization technology, ExpEther" - Extending the PCI-e bus through 10Gb ethernet. A host talks PCI-e to an FPGA which encapsulates PCI-e messages over ethernet and transmits them to a remote FPGA which decapsulates the PCI-e messages and passes them to a PCI-e endpoint. We explored similar issues with the IPOnly server. Being able to "remote" a device without needing a host (general purpose CPU) to be connected to it is pretty neat. "Efficient Fault Tolerance in Multi-media Applications through Selective Instruction Replication" - Not all instructions are created equal. Some matter to successful completion of the program, some don't. In media applications, Control flow is important, but manipulations on individual pixels less so. No point in replaying or replicating instructions which don't matter. | | Sunday, March 2nd, 2008 | | 8:35 am |
I'm in Seattle this week for ASPLOS and VEE. My hotel room has a balcony
with a view:
Yesterday I did the touristy thing and visited the space needle,
science fiction museum and other nearby attractions. The best part of
the day was the awesome Seattle Duck
Ride. Unfortunately, my camera's batteries decided enough is
enough just as we boarded the duck, so you get a picture of the space
needle instead.
The plan for today is to get some work done, visit the Seattle
Aquarium and otherwise have fun. ASPLOS begins tonight! For the curious, our ASPLOS paper is now online.
| | Sunday, February 3rd, 2008 | | 12:44 pm |
Just registered for ASPLOS '08 and VEE '08, both to take place in Seattle on the first week of March. If you'll be there, make sure to say hi! | | 10:34 am |
| | Tuesday, January 22nd, 2008 | | 10:44 am |
SIGOPS Operating Systems Review Special Issue on
Research and Developments in the Linux Kernel
The Linux kernel, since its inception in 1991, has captured the
interest of many thousands of developers and millions of users. It
recently celebrated its 16th anniversary, includes many millions of
lines of code, and is used in production systems around the world. It
is also advancing at an increasingly rapid pace, undergoing many
changes every single day.
For this OSR special issue, we welcome technical papers covering the
latest advances that have been or will soon be merged into the Linux
kernel, as well as wild idea papers discussing promising experimental
work. In recognition of the current chasm that we wish to bridge, we
encourage papers from both the Linux kernel community and the research
community. The OSR issue aims to:
a) expose members of the Linux kernel community to exploratory
research work that is going on which might influence Linux's
evolution, and
b) expose members of the systems research community to the latest
happenings in a mature, production kernel that is widely used and
advancing rapidly.
Please submit papers related to all aspects of the Linux kernel. In
particular, papers are solicited for the the following areas:
* Virtualization
* I/O, networking and interconnects
* Support for multi-core and heterogeneous CPUs
* Co-existence with other operating systems
* filesystems, clustering, SSI
* Profiling, performance tuning, debugging
* Scaling up (e.g., supercomputers) and down (e.g., embedded devices)
* Experience reports
* Research directions
Submissions should be between 5 and 10 pages (in a font no smaller
than 10 pt). Standard SIGOPS formatting rules apply (see
http://www.sigops.org/osr.html). Papers should report on significant
new results or directions and include at least some material that has
not been published before. Papers will be reviewed by the guest
editors and the review committee. Accepted papers will be published in
the July 2008 issue of Operating Systems Review.
Please upload your submissions (in PDF format only) to the submissions
website, at http://www.easychair.org/conferences/?conf=osrlk2008. If
you have any questions or comments, please do not hesitate to contact
the guest editors.
Important dates:
Submission deadline: March 14th, 2008
Author notification: April 18th, 2008
Camera-ready submission deadline: May 13th, 2008
OSR guest editors:
* Muli Ben-Yehuda, IBM Haifa Research Lab <muli@il.ibm.com>
* Eric Van Hensbergen, IBM Austin Research Lab <bergevan@us.ibm.com>
* Marc E. Fiuczynski, Princeton University <mef@cs.princeton.edu>
Review committee:
* Patrick Bridges (University of New Mexico)
* Angela Demke Brown (University of Toronto)
* Hubertus Franke (IBM Research)
* Oren Laadan (Columbia University)
* Paul McKenney (IBM Linux Technology Center)
* Chris Mason (Oracle)
* Ron Minnich (Los Alamos National Laboratory)
* Stephen C. Tweedie (Red Hat)
* Chris Wright (Red Hat)
* Pete Wyckoff (Ohio Supercomputer Center)
| | Saturday, January 5th, 2008 | | 11:36 am |
While walking the streets of Rome a month ago, I couldn't stop thinking of Bernard Cornwell's Arthur series of books. For some reason, experiencing the Roman Colosseum first hand and listening to the (excellent) tour guide tell gory tales of games and slaughter, brought the Arthur books - which I've read and re-read, but not in recent years - vividly back to mind. Now that we're back home and I've finished re-reading all three books, I still don't know why Rome reminded me of them. Which is not to say I didn't stay up until the wee hours of the morning devouring them. | | Thursday, January 3rd, 2008 | | 10:56 pm |
ASPLOS '08 program is online. Looks exciting. The Operating Systems Review special issue on Systems Work from IBM Research should be going to press any day. It features an editorial, two papers and an introduction by me, and lots of excellent papers from IBM Research and elsewhere. I am still searching for the blog voice I used to have. Will let you know when I find it. | | Sunday, December 9th, 2007 | | 11:27 pm |
Back from Rome. It's amazing how rested and ready to take on the world
I feel, after only three days.
I may write a summary later; in the mean time, as usual, Orna does it better.
| | Wednesday, December 5th, 2007 | | 9:42 pm |
We're off to Rome in a few hours for a well-needed and much-deserved vacation. | | Wednesday, November 7th, 2007 | | 7:55 am |
There's nothing quite like starting your morning, still half asleep, with a flat tire, halfway on your way to your one-year-old's daycare. I am getting proficient in replacing flat tires though - this one only took 20 mins end-to-end, despite the car being on a slope and various baby-induced distractions. | | Thursday, November 1st, 2007 | | 10:55 am |
| | Sunday, October 28th, 2007 | | 8:00 am |
SYSTOR 2007 begins tomorrow. I'm so excited I barely slept last night. I keep wondering what I forgot, what am I missing, what will go wrong. It should come as no surprise that I'm tired and cranky. But hey, on the positive side, I think I figured out at 5AM while staring at the ceiling and listening to the sleeping house a Calgary bug that had me baffled for a while. It only occurs on a new class of servers where some busses are on CalIOC2 and some aren't, a topology we haven't seen before. Two papers, an editorial and a short introduction due tomorrow for the Operating Systems Review special issue on work from IBM research, which will also have selected best papers from SYSTOR. One of the papers is nearly done, the other is making steady progress, the editorial needs some lovin' and the introduction I have yet to start. Oy. | | Thursday, October 25th, 2007 | | 1:35 pm |
hell yeah, redux
Dear [Lead Author]
Congratulations! On behalf of the ASPLOS 2008 Program Committee, I am
delighted to inform you that the your submission to ASPLOS 2008 has
been accepted to appear at the conference:
Tapping into the Fountain of CPUs - On Operating Systems Support
for Programmable Devices
The Program Committee worked very hard to thoroughly review all
submitted papers. They selected 31 papers out of 127
submissions....
Current Mood: HAPPY HAPPY JOY JOY |
[ << Previous 20 -- Next 20 >> ]
|
![[info]](http://l-stat.livejournal.com/img/userinfo.gif){kind=small}
mulix's journal