Become a fan of Slashdot on Facebook

 



Forgot your password?
typodupeerror
×
Security Businesses Google The Internet

Google Shares Its Security Secrets 106

Stony Stevenson writes "Google presents a big fat target for would-be hackers and attackers. At the RSA conference Google offered security professionals a look at its internal security systems. Scott Petry, director of Google's Enterprise and founder of security firm Postini, explained how the company handles constant pressure and scrutiny from attackers. In order to keep its products safe, Google has adopted a philosophy of 'security as a cultural value.' The program includes mandatory security training for developers, a set of in-house security libraries, and code reviews by both Google developers and outside security researchers."
This discussion has been archived. No new comments can be posted.

Google Shares Its Security Secrets

Comments Filter:
  • Security secrets? (Score:5, Informative)

    by illegibledotorg ( 1123239 ) on Thursday April 10, 2008 @12:46PM (#23026252)
    TFA is a little scant on "security secrets."

    What is covered is some general security policy and philosophy.

    And here I was, waiting to read all about GIDS and GFirewall. Thanks, ITNews, for instead educating be about archiving security logs for later review!
  • Fluff Acticle (Score:1, Informative)

    by Anonymous Coward on Thursday April 10, 2008 @12:50PM (#23026300)
    How does an article that has no technical content, no news, and no information make it to the front page of a tech new site? Oh yeah, this is Slashdot, fake journalism at its best.
  • by kevin_conaway ( 585204 ) on Thursday April 10, 2008 @12:51PM (#23026324) Homepage

    Tools like PMD help with this .

    We ended up getting bitten by bugs like unsynchronized access to static DateFormat object so we wrote used a PMD rule to fail our build if anyone does such a thing. We have other rules that curb the use of IOUtils.copy (instead of copyLarge).

    I highly recommend using some sort of static analysis as part of your CI process

  • by morgan_greywolf ( 835522 ) * on Thursday April 10, 2008 @01:20PM (#23026740) Homepage Journal

    I'd be surprised if any from-scratch operating system designed for internet-facing use today, didn't also have 'security as a culture'.
    Yeah. It's called OpenBSD [openbsd.org].

  • by Shados ( 741919 ) on Thursday April 10, 2008 @01:38PM (#23026998)
    What the previous poster was refering to is that serious development shops will use code analysis tools to inforce it: that is, the code will not be allowed to be checked in (or to be integrated to the trunk, or whatever) if the rules are not followed, and they are inforced at the source control level (or something).

    Variations include having the code analysis tool throw "compiler" warnings, and make the compilation to consider warnings as errors and fail the build.

    Once you start working in an environment that does such things, you don't go back: the code quality goes up 10x.
  • physical security (Score:3, Informative)

    by Kartoffel ( 30238 ) on Thursday April 10, 2008 @02:18PM (#23027662)
    What about physical security for Google facilities? Last time I was in Mountain View I took a leisurely stroll right through the middle of the Googleplex, right past the life sized dinosaur skeleton, right past the sand volleyball court and hot tub and right through a couple of their office buildings. I like how the Googleplex is set up like an academic campus, but it's pretty trivial for a bad guy to bypass the card access doors by piggybacking behind somebody else.

    Also, the whole place is made out of floor to ceiling glass windows. Would be really simple to shoulder surf somebody's display through a telescopic lens or listen against a windows with a laser mic. There's a reason high security buildings tend to resemble windowless block houses. Hopefully, anybody with a window seat at the Googleplex never processes sensitive data.
  • by Medievalist ( 16032 ) on Thursday April 10, 2008 @05:43PM (#23030040)

    C'mon, I know you do it too: when I want to see if my Internet is working, I "ping www.google.com".
    I still find it surprising that it ICMP_ECHO_REPLYs my ICMP_ECHO_REQUESTs. Why?
    I find it surprising that you find it surprising! :)

    A lot of sites disable ping because, years ago, The Ping of Death could crash a server by sending maliciously-crafted ping packets.
    The "Ping of Death" gained fame because any chump could create one from a totally generic Windows system using the broken ping that Microsoft was shipping at the time. The technique is applicable to any IP protocol, not just ICMP echo. You can make an SMTP of Death fairly trivially. Just fake up a datagram with a total length greater than 65,535 by abusing the fragment offset field of the IP header, and if the target system does not check total length for validity you can overflow memory and hose the system. If that didn't make sense to you, just remember the "Ping of Death" has NOTHING TO DO WITH PING - it's an IP vulnerability that used to exist for ALL protocols in the IP stacks of certain vendors (IBM, Sun, Cisco, etc.) and is now fixed.

    And you can DOS a server by flooding it with pings.
    And you can do it more easily with practically any other type of packet. If you plan to block all traffic that can be used for DOS, you must block all traffic, period.

    Ping is a service we all should provide to our internal networks from individual hosts, and to the Internet at large at the network edge. Configure your routers to respond to pings for your hosts instead of passing them through the firewalls. Ping is how people who need to test their ability to reach your hosts or site can do so. It is a simple tool that consumes a minimal amount of bandwidth to get the job done.

    I'd be interested to know just how many pings Google receives, and replies to each day.
    They might tell you if you ask. If it ever gets out of hand they'll just respond with normal traffic shaping techniques.

    And how many of those are maliciously encoded, only to be defeated by the ub3rh4x0r5 at Google.
    There's nothing dangerous about ping. Nothing... you can tell if a network is competently administered just by pinging it, my friend. I'd never hire anyone who had an unpingable net.

    Hmmm... where's BadAnalogyGuy when you need him? OK, look, blocking ping is like saying that you've seen a guy killed by an Isuzu truck, so you think you can prevent all fatal accidents by banning Isuzu trucks from the highway. In reality, all you will do is prevent beer deliveries to my house, since my beer distributor uses Isuzus. This will make me hate you, just like people hate clueless firewall admins who block ICMP. Or wait, you saw a guy get bludgeoned to death with a hammer so you will ban all hammers while allowing people with large wrenches, razor knives and screwdrivers to pass without comment. That was pretty bad I think.

BLISS is ignorance.

Working...