Plumbing Master, Plumbing Apprentice
June 28, 2009
Thank goodness for Cade Metz’s “Google Mocks Bing and the Stuff behind It.” My research documented Google’s investment in hardware and software infrastructure once Messrs. Bing and Page had the cash to move away from the Lego rack housing its first servers. Microsoft’s approach to infrastructure has been via a more traditional approach. I have described the basics of the company’s approach in this Web log.
Microsoft’s cutting back on its data center spending underscored for me the burden Microsoft’s approach imposed on the company in a faltering economy. Hiring Yahoo data center experts struck me as a clear indication that Microsoft wanted to follow a different drummer than Microsoft.
I routinely include diagrams of Google’s approach in my lectures. No one in my audiences in organizations at gatherings like the NFAIS venue on June 26 seems to know that hardware and software plumbing is a Google core competency.
Mr. Metz’s write up observed:
“Our approach is a little more absolute than [Microsoft’s],” Gill said. “Not only does getting to the end user have to be fast, but the back-end has to be extremely fast too…[We are] virtualizing the entire fabric so you get maximum utilization and speed on a global basis as opposed to local fixes – putting one service in a data center, for example, in Denver. “You want to figure out how you want to distribute that across the entire system so you get it as horizontal as needed, which is essentially the definition of cloud computing.” You can take issue with his terminology. But his argument is sound: While Microsoft is struggling to separately hone performance for each and every application, Google can uniformly juice speed across its entire portfolio. The secret to Google’s success, Gill said, is not in the company’s mystery data centers, but in its software infrastructure, including GFS, its distributed file system; BigTable, its distributed database; and MapReduce, its distributed number-crunching platform.
Several points warrant mentioning:
- Google’s approach delivers costs savings in a number of ways, including the automation of data center operations, configuration, spending for uninterruptable power supplies, etc.
- The Google plumbing is more homogeneous than Microsoft’s. You can see what I mean by looking at the diagram in my write up here.
- Google has done a lousy job of explaining its advantage. Content to publish technical papers, the language, the Googley humor (janitors for the smartest software in the Googleplex), and the game plan that Googlers run when giving talks give Microsoft ample scope to market their brand of technical excellence.
The bottomline is that Google has bumbled the opportunity to make clear what it has built and what its “as is” infrastructure can do.
I am offering my opinion based on Mr. Metz’s write up. In my Google studies, I tried to be clear and explain what Google achieved. I don’t think anyone really understands Google’s achievement and using Googley humor to make the point works in Microsoft’s favor. Math club is fun for those in it. Those who are in other activities don’t know what the heck the math folks are talking about.
Big, big problem for Google.
Stephen Arnold, June 28, 2009
Comments
8 Responses to “Plumbing Master, Plumbing Apprentice”
-
- Subscribe to Beyond Search
Feature archive
News archive
-
I disagree. On the contrary, I’ve run into a number of people who think MapReduce has magical powers. The advantage of Google infrastructure is large but not insurmountable.
Unfortunately, the Register article did not do a very good job of covering Microsoft’s new and substantial efforts at building Google-like distributed primitives like Azure, SSDS, and Axum.
If you think Google has done a poor job trumpeting their infrastructure assets in the form of MapReduce, GFS, BigTable, Sawzall, Chubby, etc then you probably are moving in the wrong circles.
There are now lots of venture funded startups either trumpeting or using knock offs of Google’s infrastructure assets (i.e. Hadoop) so it seems crazy to hear someone argue that they haven’t done a great job of getting the word out.
Hi Stephen,
Interested to hear your assessment of the NFAIS attendees (I’m not so surprised at that – they aren’t techie types generally). You seem to think that it is a weakness of Google that they don’t shout about their clear strengths in Hardware/Software. I have thought for a long time now that this has been a very deliberate approach. When Marissa Mayer (if memory serves) commented in an article about how the 401 phone service (I’m in the UK, but that’s directory services right?) was really about getting enough data to do voice recognition for search based activities, I thought “Hello, these people are operating on a different level to everyone else.”
Question – would Googles powers with data, scare people if they understood it? I think the answer there could all too easily be Yes. That would be bad for shareholder value as scared people probably ring their politicians and ask for things to be done about it.
[…] Plumbing Master, Plumbing Apprentice : Beyond Search […]
[…] Plumbing Master, Plumbing Apprentice : Beyond Search […]
[…] Plumbing Master, Plumbing Apprentice : Beyond Search […]
[…] Plumbing Master, Plumbing Apprentice : Beyond Search […]
You misspelled Sergey’s last name as “Messrs. Bing and Page” in paragraph 1. Unintentionally, I’m sure.