Forgot your password?
typodupeerror
Government Security The Internet IT

80% of .gov Web Sites Miss DNSSEC Deadline 79

Posted by kdawson
from the not-a-priority dept.
netbuzz writes "Eighty percent of US federal agencies — including the Department of Homeland Security — have missed a deadline to deploy DNS Security Extensions, a new authentication mechanism designed to prevent hackers from hijacking Web traffic. The deadline that whooshed by was Dec. 31, 2009. Experts disagree as to whether this level of deployment represents a failure or reasonable progress toward meeting a mandate set by the Office of Management and Budget in the summer of 2008. OMB officials declined to say why the agency hasn't enforced the DNSSEC deadline for executive branch departments."
This discussion has been archived. No new comments can be posted.

80% of .gov Web Sites Miss DNSSEC Deadline

Comments Filter:
  • I'm not a huge fan of DHS, but come on there are so many other government agencies that hardly ever get any abuse at all. DHS has had a lot of cock ups, and should be ridden hard to shape up or dissolve, but this is hardly an opening sentence kind of cock up.

    Now where is the full list of orgs that have or have not done it? I suspect its going to be a lot like reading the pork report.

    • by Sir_Lewk (967686) <sirlewk@nOSpAm.gmail.com> on Friday January 22, 2010 @12:22PM (#30860458)

      The reason why the DHS gets more attention here than other departments is because they are the Department of Homeland Security. The importance of irony when ridiculing the government is not to be overlooked.

    • Re: (Score:3, Insightful)

      by Opportunist (166417)

      Now where is the full list of orgs that have or have not done it?

      Why, looking for a shopping list? :)

      Seriously, this time I could even understand if it was not released for "reasons of national security". It would be one of the few cases where that excuse actually makes sense.

      • Re: (Score:3, Insightful)

        by tiberus (258517)
        First, let's hope it's a reason and not an excuse...
        Second, Security through obscurity is no security at all or No security through obscurity.
      • Re: (Score:3, Insightful)

        by Smallpond (221300)

        Seriously, this time I could even understand if it was not released for "reasons of national security". It would be one of the few cases where that excuse actually makes sense.

         
        Because the terrorists who are going to attack using a sophisticated DNS cache poisoning technique are obviously too stupid to download a list of government websites and go through them one-by-one to see which are using DNSSEC.

    • Re: (Score:2, Funny)

      by Anonymous Coward

      Sir, I think that you have penis on the brain.

      And I quote:
      I'm not a huge fan of DHS, but come on there are so many other government agencies that hardly ever get any abuse at all. DHS has had a lot of cock ups, and should be ridden hard to shape up or dissolve, but this is hardly an opening sentence kind of cock up.

      Now where is the full list of orgs that have or have not done it? I suspect its going to be a lot like reading the pork report.

    • Re: (Score:2, Interesting)

      by Zibri (1063838)

      Seems that most of the larger (well-known) *.govs doens't haven't deployed dnssec. I tried cia.gov, fbi.gov, nsa.gov (!), state.gov, whitehouse.gov, ins.gov, irs.gov... state.gov was the only one i found having published a DNSKEY rr. (I just picked a few at random I knew)

    • by mcgrew (92797) *

      IMO "Homeland Security" should have never been established in the first place, and once established FEMA should not have been part of it.

      The armed forces are supposed to secure the homeland.

      I take issue with the name itself as well; "homeland" puts pictures of Nazi Germany in my head. Maybe thay did that on purpose?

      • by lorenlal (164133)

        I take issue with the name itself as well; "homeland" puts pictures of Nazi Germany in my head. Maybe they did that on purpose?

        Normally, it's customary to at least wait a *little* while before dropping that line. [wikipedia.org]

        • by mcgrew (92797) *

          I knew someone would mention Godwin, but sometimes reference to the Nazis is justified.

      • by e9th (652576)
        Who comes up with names like these? "Homeland" is disturbingly close to "Vaterland." Wouldn't "domestic" have worked?
        My county renamed the Sheriff's Office to "Department of Public Safety," bringing to mind that laff-riot Reign of Terror during the French Revolution.

        And yes, DHS could easily have been made part of the FBI or the US Marshall's Service, if it needs to exist at all.
      • I take issue with the name itself as well; "homeland" puts pictures of Nazi Germany in my head. Maybe thay did that on purpose?

        I'm not going to Godwin you, but you do get a citation for 'ignoring a common cause'. The US has increasingly trended towards fascism (the actual political system, not the angsty high school usage).

        • by mcgrew (92797) *

          The US has increasingly trended towards fascism

          Yes, it has, and I hate the fact that it has. I wish I could see a way to stop it.

  • by brennz (715237) on Friday January 22, 2010 @12:22PM (#30860456)

    (1) you have a shill of a biased company selling products to the industry pushing the requirement
    (2) An unrealistic deadline set by OMB initially.

    This is a craptastic story.

    • Re:of course (Score:4, Insightful)

      by Archangel Michael (180766) on Friday January 22, 2010 @12:59PM (#30860922) Journal

      1) Yeah? And?

      2) IT wasn't unrealistic.

      How long does it take to implement?

      1) Get deadline
      2) Start product evaluations
      3) Pick Product(s)
      4) Implement Product
      5) Write Howto: for all the idiots out there

      If we use 3 Months (1/4 year) for each step, we're looking at 1 year, three months to implement, including figuring out time lines for implementation.

      Once you start rolling out, you cookie cutter as much as you can, so you have easy, consistent configurations and implementations.

      I don't get why it takes so long.

      • Your Concept of project management is akin to a PHB on software development.

        Your program need the following

        1) Take Input from User
        2) Process Input
        3) Save Information
        4) give output.
        5) Document for future use.

        If we use 3 Months (1/4 year) for each step, we're looking at 1 year, three months to implement, including figuring out time lines for implementation.

        Once you start rolling out, you cookie cutter as much as you can, so you have easy, consistent configurations and implementations.

        I don't get why it takes

        • Re: (Score:3, Insightful)

          DNSEC is not the same as writing a program. It is a service that does one thing. DNS .. Securely.

          The protocols are ALREADY set, it is just a matter of configuration and implementation.

          Again, other places have DNSEC working right, so what is so hard about getting it working here? I mean besides normal Bureaucratic Government Ineptitude?

          • by Lennie (16154)
            You have no idea. DNSSEC takes a lot of time, it's complicated, many things can go wrong. If just one thing goes wrong it completely fails. On top of that it needs to be updated regularly.

            That's why DNSSEC hasn't been implemented by 80% of ISP in the world. Like the 'fix' for the Kaminsky-attack.
      • Re: (Score:2, Insightful)

        by Anonymous Coward

        If we use 3 Months (1/4 year) for each step

        You've never worked for a government agency, or on a government contract, have you? 3 Months for "Getting the deadline" is usually unrealistic!

        NOTE: This reality pisses me off to no end...

        • by spamking (967666)
          Most deadlines in the government are doomed from the beginning . . . there is so much that most projects have to go through just to get considered not counting the implementation phase it is ridiculous. Until folks have actually worked within the government they'll never understand.
          • Re: (Score:1, Insightful)

            by Anonymous Coward

            That's because we elect/hire people into government positions that:

            1) look good
            2) are well spoken (talk a load of shit, but do it well)
            3) have the majority of their education devoted to playing the system (lawyers, MBAs, etc.)
            4) are endorsed by moneyed interests

            rather than people that:

            1) Look normal
            2) Act dumb on camera but get shit done
            3) Have the majority of their education devoted to their field of expertise (Doctors, Engineers, Climatologists, etc.)
            4) are endorsed only through public funds.

            When you put

      • by brennz (715237)

        you left out anything about budget, or acquisition activities.

        If you think the govt has good IT people and loads of $$ just sitting around waiting for unfunded mandates from OMB, you are smoking something.

        http://en.wikipedia.org/wiki/Parkinson's_Law [wikipedia.org] applies to govt IT and $$

    • Re: (Score:3, Interesting)

      by Sir_Lewk (967686)

      I can certainly understand the unreasonable deadline complaint, but why exactly is DNSSEC "just some product being pushed by a shill company"? BIND implements DNSSEC, it's not like it's a proprietary piece of technology that is only offered by a single vendor.

      • Does Bind have a point-n-click gui so the MCSE's can use it? One can't expect them to edit text files to configure the software after all.

      • by jafiwam (310805)

        A lot of internal DNS is done with a Windows Domain Controller and the built in DNS there.

        So it's not enough that BIND does it, Windows servers need to as well.

        Next, consider the number of Windows2000 networks still out there and the problem of implementation becomes more tangled.

        I am using BIND, but not DNSSEC. The GUI that sits in front of mine does not have any options for DNSSEC.

        Just go look at the Wikipedia article on DNSSEC and the thick mass of references and new terms and it's FUCKING OBVIOUS it's

        • by Sir_Lewk (967686)

          I'm not arguing that it should be easy to do, or that their deadline was reasonable. I've worked with BIND before and fully acknowledge that it's a bitch and a half. What I'm asking about is why the original poster seems think that DNSSEC is a worthless product being pushed by a single company looking to make a buck.

          • by Lennie (16154)
            Because ISC is one of those companies that is pushing DNSSEC and a lof ot other DNS-companies are not, because they think it's a bad idea to implement something which is so complicated. Many security bugs arrise when things start to get complicated.
    • I'm pretty sure more money has gone into lobbying against DNSSEC than in favour ... it's going to have a really big toll on the CAs after all when everyone can just put a self-signed cert inside their DNS entry and have end to end authentication completely without a CA.

      • by Lennie (16154)
        I really wish we can get browser (and other client)-support for this soon, that would be such an improvement.

        The only 'applications' we currently have that supports fingerprints in DNS are some implementaitions of IPSEC and SSH.

        Even Dan Kaminsky would probably agree to that. Especially if it wasn't based on ASN.1 like current SSL-certs.

        I really hate the the whole structure of how the whole CA-business work and how SSL-certs are constructed, it's a big mess.

        StartSSL is OK, CACert would have been even better
  • by Monkeedude1212 (1560403) on Friday January 22, 2010 @12:27PM (#30860544) Journal

    Rumour has it All Canadian governments open TCP/UDP ports 2 through 65535.

    The first one is the reserved emergency port for the Prime Minister to escape in the case of a national emergency. We tried to explain to him that's not how it works but... You know politicians...

    • Re: (Score:1, Funny)

      by Anonymous Coward

      It's called the "Diefenportnumber".

  • by Anonymous Coward

    They'll do a much better job when they gatekeep everyone's health records.

  • by adosch (1397357)
    This is probably more of a classic case of unrealistic deadlines imposed on Gov't agencies/IT contractors by Gov't security desk jockies and/or congressmen without a clue. I'm sure the infrastructure is convoluted to begin with and I'm sure whatever planning testing was probably rushed. On top of that, I've never know *anything* in the government to 1) rarely meet a deadline on time, 2) accomplish a task on time without an exorbitant amount of hiccups to deal with, or 3) be successful without being stran
  • by King_TJ (85913) on Friday January 22, 2010 @12:36PM (#30860650) Journal

    Sure, it's always good to implement updates that improve network/computer security ... but let's face it. These deadlines are put in place primarily to ensure people actually pay attention and do the update in a reasonable amount of time. It's not like govt. had inside information that right after Dec. 31, 2009 - hackers were going to go crazy trying to exploit this DNS issue, so that was the day it really NEEDED to be implemented by, across the board.

    Maybe I'm just in a sour mood right now with this stuff in general? But lately, I sense an ever-increasing amount of importance being placed on every little security patch or change, when it's just not really warranted. It seems really self-serving to those who work in the field of "computer security", because it makes a bunch of extra billable work for them - and they get to scare more people into paying them to secure things for them.

    I mean, just this morning, I came into work and checked my mail, and what do I see? People on C-Net asking questions about if they should just "quit using Internet Explorer, given the recent security exploits". (Umm, let's see here.... You successfully used the thing ever since probably when? At least back in 2001 or 2002, right? And theoretically at least, it's "safer" now than EVER before, since Microsoft has been patching and upgrading the thing that whole time. So why would you suddenly determine NOW that it's just too unsafe to use again??)

    And later today, I've got to waste my afternoon ensuring "PCI Compliance" because my workplace accepts credit cards once in a while, processed via an Internet-based card processing service. We don't even store *any* of the card data here, on either our systems or on paper. They just punch the stuff into the web site to do the processing, and let the processor keep the data. But *still*, simply because we do it, we have to have monthly "penetration testing" done against our firewall's IP address (among other requirements), and the stupid test claims I "fail" right now, due to issues that hardly matter in reality. (EG. It's complaining about unpatched issues with the Outlook Web Access part of Exchange, even though nobody even has access to use OWA in our company except me, as sysadmin -- and again, I'm finding it quite the stretch to see how someone hacking OWA here would magically obtain customer credit card info, given how we operate here?)

    • by Mashdar (876825)
      Maybe not you, but what about the guy down the road? Making special exceptions just because you do X Y or Z, or don't do A B or C does not make sense. Good rules always err towards over-protection. Not to mention the fact that the information is probably cached locally, and how does anyone outside your business know what you do and do not do with OWA or anything else? Or what you might do in three months when you decide to switch services.
  • Just out of curiosity, I would love to check if my state is compliant.. How does one use NSLookup or DIG to check?? is it just a txt field, like looking for an SPF key?

    Of course, to fully check I would have to check the keys from .Gov, then the key from the domain, so do either tool have the capability to "walk the tree"?

  • by RichMan (8097) on Friday January 22, 2010 @12:52PM (#30860826)

    So does this show a lack of government IT ability. Or is it more representative of the general inertia of government. I would worry more about the former. Where the government is exposing itself to the wilds of the internet without the ability to protect itself.

  • Good... (Score:5, Interesting)

    by nweaver (113078) on Friday January 22, 2010 @12:56PM (#30860868) Homepage

    DNSSEC still has some serious problems. EG, in our preliminary analysis, a shockingly large number of Netalyzr users are behind DNS resolvers that can't handle fragmented traffic. Yet a large number are behind resolvers that do request DNSSEC data.

    Since DNSSEC replies are often large (and can easily be over the 1500B response limit), turning on DNSSEC could very well mysteriously slow down DNS by causing large timeouts as the UDP reply fails to arrive and the DNS resolver, after a long timeout, then resorts to a TCP connection, even when the signatures are not validated, simply because there are a lot of resolvers that request DNSSEC but actually can't handle large replies.

    http://www.ops.ietf.org/lists/namedroppers/namedroppers.2009/msg01513.html [ietf.org]

    • DJB's DNS curve would have solved this problem.

      • by nweaver (113078)

        No it doesn't.

        The big deal is DNSCurve doesn't solve the real threat:

        DNSCurve provides transport integrity against in-path adversaries OTHER than the recursive resolver. DNSSEC provides protection against malicious recursive resolvers.

        But transport integrity for DNS is pointless, as anyone in-path on your DNS data is also in-path on the rest of your data.

        As important, DNSCurve does NOT protect against the resolver misbehaving, which we have witnessed on multiple occasions (some ISP's will MitM google by re

        • by Cyberax (705495)

          Yes, it does.

          DNS curve is designed to use small UDP packets. And it's more secure because it encrypts the packets' contents. But I guess that deep packet inspection folks won't allow that.

          DNSSEC doesn't protect against recursive resolvers. I can set up a malicious resolver at my ISP which will just strip the DNSSEC records, not a problem. End user software still must validate the signatures.

      • DNSCurve is an elegant and efficient idea, and most importantly, it would be easily deployable--requiring minimal changes to infrastructure, and no cooperation from end users or support in end devices. The only obstacle that I am aware of, is that no implementation exists. (Which is a problem...)

        It does not provide end-end trust, but it is close enough in most cases, and you can always run your own local DNSCurve forwarder if you need the extra guarantee.

        Even so, DNSCurve and DNSSEC are complimentary solu

    • by Tacvek (948259)

      Off Topic: I just tried the netalyzr with some interesting results. When I run it under Firefox 3.6 I get terrible results for HTTP caching. Of note is that the "diretly and explicitly" request tests for strongly and weekly uncachable data fail. When I run it under Chrome I get the expected results of no indication of an HTTP cache. So I suspect Firefox may be interfering with the HTTP caching tests.

      I would like to suggest another port test, for port 6667, the default port for Internet Relay Chat. Since man

      • by nweaver (113078)

        REALLY odd, Java is not supposed to do that because we are directly connecting to port-80. Could you contact netalyzr-help@icsi and send us the URLs for the results pages?

  • by Anonymous Coward on Friday January 22, 2010 @01:07PM (#30861022)

    I am the DNS admin of a federal agency. We signed two of our domains, and twice had .gov delete the keys that allowed the domains to be trusted. We then got the run-around and were lied to by the .gov admin. My management and I are now afraid to make any further progress implementing DNSSEC because .gov has made so many mistakes. It is better to be unsigned than to be signed and have the trust keys be incorrect.

    Additionally, the tools to implement DNSSEC are non-trivial. A federal agency or Fortune 500 can afford to buy a Secure64 Signer. Looking forward to when I want to sign my personal domains (in .org and .com), the tools have to become much simpler and much more automated.

  • by snsh (968808) on Friday January 22, 2010 @01:10PM (#30861078)
    I manage a .gov domain for a non-federal entity. Last year I pursued DNSSEC and hosted DNS to improve availability and diversity over our on-premise DNS. Windows DNS and BIND seemed okay for DNSSEC secondaries, but signing and key rollover are high-maintenance. Maybe in the near future that will change. There are appliances I could buy for $10-20k to manage master zones and do DNSSEC, but they were out of budget. I worked with a hosted provider (dynect) for DNSSEC singing with .GOV, but that turned out to be out of budget too. So eventually I just settled on dnsmadeeasy for nominal cost, with anticipation that they'll support DNSSEC sometime in mid-2010. Basically DNSSEC for the masses doesn't seem to be there yet.
    • but signing and key rollover are high-maintenance

      Could you clarify the problems you faced here? I've not found those to be such high hurdles with BIND, but our experiences probably differ.

  • IMHO, the reason this isn't done yet is because of the org structure. OMB is responsible for administrative oversight of this type of stuff, but each department don't actually work for them obviously.
    So it could be analagous to the corporate IT department sending an email to each department lead (sales, production) telling them to install certain patches to their desktop PC.

    Yeah sure, the IT department has the right to give direction because the common CEO delegated that responsibility to them, but when pr

  • ... DNS Security Extensions, a new authentication mechanism designed to prevent hackers from hijacking Web traffic.

    "New"? From: Domain Name System Security Extensions [wikipedia.org]:

    The initial RFC 2065 was published by the IETF in 1997, and initial attempts to implement that specification led to a revised (and believed fully workable) specification in 1999 as IETF RFC 2535. Plans were made to deploy DNSSEC based on RFC 2535.

    Oh well, "netbuzz" and KDawson are probably too young to know any better :-)

    • The Feds spent the 90s trying to block public use of encryption anywhere in the world, but especially in the US. The excuse they used was that it would weaken their ability to eavesdrop on Commies (not that there was still a Soviet Union around by then), but they were able to interfere with the development and release of a lot of open-source software by claiming that releasing it on a public server would be export and therefore covered by the ITAR munitions export laws. They even withheld approval for "bo

  • I bet 80% of .gov sites also don't have properly setup DNS, let alone DNSSEC.

    eg. Try going to http://fbi.gov [fbi.gov]

    Without the proper CNAME record you need to type "www" before the hostname. Silly.

  • This is what the subject line in my RSS reader (Thunderbird) just gave me:

    4 Out of 5 of<nobr> <wbr></nobr>.gov Web Sites Miss DNSSEC Deadline

    WTF? Are you writing this stuff in MS Word?
    Because I constantly see this stupid shit. And no human would ever do something like that.

  • by Lennie (16154)
    We need DNSSEC on the root, w00t, w00t! ;-)

    No really, without DNSSEC on the root, I don't think we'll get proper verification process going on the resolver side.

    And putting something in DNS which isn't verified is hardly useful. Maybe they will do verification within the government, that is a start.

The trouble with opportunity is that it always comes disguised as hard work. -- Herbert V. Prochnow

Working...