Recent Updates RSS Hide threads | Keyboard Shortcuts

  • Obama Picks Aneesh Chopra to be Federal CTO

    peter 11:47 pm on April 18, 2009 | 0 Permalink | Reply

    Another practical move by the Obama administration. O’Rielly loves this pick. I’m not so sure. Personally I would have liked to see a CTO from the private sector get in there and shake things up.

     
  • Obama's CIO Pick Takes Leave of Absence

    peter 10:29 am on March 14, 2009 | 1 Permalink | Reply

    Just days after a Federal raid on his former office, Washington Post is reporting that Vivek Kundra has now taken a leave of absence

    Of all the CIO/CTO’s in all the land I’m still not sure why Obama picked a CIO from the public sector which is always way behind the curve. And now there seem to be a scandal. Seems like a giant missed opportunity to me.

     
    • mr.sears-parts 9:01 pm on April 19, 2009 Permalink

      Hello, I can’t understand how to add your blog in my rss reader.plz tell me thank you.

  • Obama's CIO

    peter 9:13 pm on November 20, 2008 | 0 Permalink | Reply

    We are all waiting patiently for Obama’s CTO pick. Now there is even a web site for it.  I can’t wait to see a savvy CTO get in there and clean this bloated government sector up. It’s past time to start bringing public service into the Internet age.

     
  • Low Cost Infrastructure - A Look at Wikipedia

    mat 6:23 am on June 25, 2008 | 1 Permalink | Reply
    Tags: Domas Mituzas, infrastructure, linux virtual server, , O'Reilly Velocity conference, wikipedia

    A piece at Data Center Knowledge about Wikipedia’s infrastructure. Being a non-profit with no revenue at risk and small budgets to spend, they take a very different approach to availability, running one main co-lo in Florida with about 300 servers and focusing on performance more than availability.

    “The traditional approach to availability isn’t exactly our way,” said Mituzas, who spoke about Wikipedia’s infrastructure Monday at the O’Reilly Velocity conference. “I’m not suggesting you should follow how we do it. But losing a few seconds of changes doesn’t destroy our business. As long as a crash doesn’t turn into a disaster, there’s no witch hunting or heads rolling.”

    Read more here.

     
    • Richard 3:43 am on September 22, 2008 Permalink

      Where have all the CTOs gone? Is this blog dead now?
      R.

  • Google spills the beans

    mat 2:26 pm on June 23, 2008 | 0 Permalink | Reply
    Tags: co-lo, gfs, , io2008, james hamilton

    James Hamilton writes from Google IO and reports for what I think is the first time on the actual numbers of Google servers and other juicy tidbits. The full report can be found here but here’s a summary:

    1. Single search query touches 700 to up to 1k machines in 800K servers
    2. Install 1000 machines and in a year you’ll see:
      • · 1000+ HD failures
      • · 20 mini switch failures
      • · 5 full switch failures
      • · 1 PDU failure
      • · 0.5 co-lo failures
    3. >200 Google File System clusters
    4. The largest BigTable instance manages about 6 petabytes of data spread across thousands of machines
    5. MapReduce is increasing used within Google:
      • · 29,000 jobs in August 2004 and 2.2 million in September 2007
      • · Average time to complete a job has dropped from 634 seconds to 395 seconds
      • · Output of MapReduce tasks has risen from 193 terabytes to 14,018 terabytes
      • · Typical day will run about 100,000 MapReduce jobs
      • · each occupies about 400 servers
      • · takes about 5 to 10 minutes to finish
     
  • Networks and Massively Parallel Processing Environments

    mat 8:59 am on June 23, 2008 | 0 Permalink | Reply
    Tags: aster data, aster data systems, david cheriton, mpp, stanford

    Interesting article here by David Cheriton, head of Distributed Systems at Stanford. David has an interesting history as an angel investor, being an angel only three times: Google, VMWare and Aster Data Systems. This article is a neat summary of why the future of MPP is in smart software and not complex networks.

     
  • FreeBSD 7.0

    mat 11:02 am on April 2, 2008 | 0 Permalink | Reply
    Tags: freebsd, kris kennaway, linux, ,

    I’ve been a long time admirer of FreeBSD. It’s reliable and seems to focus on the things I find important for running a business than the sugar-coated desktop eye candy that many other distros spend their time on. The latest version (7.0) was launched a few weeks ago and it’s stuffed full of some really interesting features.

    Highlights:
    * End of a 7 year effort (SMPng) to make the kernel totally multi-threaded
    * Significant performance boosts for Postgres (5300 tps versus 1500 tps in 6.x and 3500 tps in Linux 2.6.23) and MySQL (3500 tps versus 1000tps in 6.x and 3000 tps in Linux 2.6.23)
    * New ULE Scheduler
    * 18,000 changes

    Read all about it in this presentation (pdf required) from Kris Kennaway.

     
  • Where have all the good databases gone?

    mat 10:50 am on April 2, 2008 | 0 Permalink | Reply
    Tags: adam bosworth, databases, , , oracle, , scalable infrastructure

    Adam Bosworth writes an insightful piece about how databases have failed to keep up with the times. I didn’t realize how much we’ve learnt to put up with things like Oracle, Postgres and even mySQL until I read the article.

    The scary thing is he wrote this article over three years ago…

    Read the article here.

     
  • good idea

    glemak 6:25 pm on March 11, 2008 | 1 Permalink | Reply
    Tags: media cto

    a group blog by cto’s – had to join in :)

    i do lots of blogging and other social stuff online right now but there’s always room for more…

    my about page has my background, which includes a stint as cto of time warner & a few other media companies…

    currently i’m in a cto-type role for hearst, another media company, focused on technology across the enterprise and our strategic investment in startups and acquisitions of online properties…

     
    • peter 11:10 pm on March 11, 2008 Permalink

      Welcome to the blog! Looking forward to your posts.

  • Microsoft & Yahoo - changing the stack in Silicon Valley?

    peter 7:15 pm on March 2, 2008 | 0 Permalink | Reply

    It is going to be interesting to see what becomes of Yahoo! in the next few weeks and months.  Will Microsoft gobble it up? Or will the dotcom elder fight off the takeover?One of the most interesting aspects of this for me is thinking about what’s going to happen to Silicon Valley should Microsoft succeed in its takeover?Owning Yahoo would give Microsoft a huge presence in Silicon Valley that the company has never really enjoyed.  Sure they’ve had a research and sales outpost here for a long time, but not a giant campus or any operating units. Microsoft would go from being a relative visitor to one of the largest employers overnight. That’s going to bring some serious changes to the Valley.One of the those changes could be make up of Silicon Valley’s engineering workforce.I believe this will center on whether or not Microsoft switches Yahoo! from a UNIX and PHP shop to a .Net infrastructure. If what happened to Hotmail is any indication of things to come, I would think some major porting is going to be on the agenda. That would mean that Microsoft would have to employ a lot more .Net engineers here in the valley. Silicon Valley is much more of a OSS/UNIX/JAVA/PHP community so I’m not sure where they would get the talent from.In fact, that could be an achilles heal of this bizarre marriage. Will it mean more .Net adoption in the valley or will Microsoft have to ship engineering jobs back to Redmond? Microsoft has recently pledged support for PHP on windows. Maybe they’ll ditch .Net all together and develop for PHP (or buy Zend)…Regardless of the stack changing, I have a feeling that Silicon Valley will never be the same. 

     
c
compose new post
j
next post/next comment
k
previous post/previous comment
r
reply
e
edit
o
show/hide comments
t
go to top
esc
cancel