The DDoS Attacks, One Year Later
ATKeiper writes: "One year after the DDoS attacks against major Web sites, C|Net reports that there are still 'no strong defenses deployed' against such attacks. The only person so far accused by prosecutors is Canadian teen hacker mafiaboy, whose trial starts in a month. Was it a forgettable stunt? A much-needed wake-up call for insecure e-commerce sites? Lame script kiddies giving hackers a bad name?"
Please form an orderly line and exit the website via the clearly marked escape routes.
I work for an ISP in the UK. We're pretty high profile here, so I can't exactly name the company. I can make some observations here though, as this topic really hits quite close to home for me.
Since December last year, we were on the wrong end of some SERIOUSLY large DoS attacks. Some of them were your run-of-the-mill smurf, but the most common has lately been a little SYN flooder which I won't mention here, lest the wannabies all go download it and try and take down Yahoo with their 56k modems. (Not that you could, you'd need more that that).
We use BTnet as our uplink provider, and initially we got very poor response from them. One attack which crippled us for 12 hours, however, managed to get their attention. Apart from the fact it wiped us from the face of the planet, stopping millions of users from dialling up or accessing their web-pages, they also managed to take out a huge chunk of BTnet's core infrastructure. BT are not happy, and neither are their customers. Strangely enough, BT has transformed into the most impressive anti-packetkiddie juggernaut I have ever seen.
Sure, it's hard to track them down, but we're learning a lot. I guess the packetkiddies think this is a one-way process. They attack and sites go down, and they think they can just keep doing it without anything happening.
Everything is in their favour, for the moment, but every single attack the packetkiddies do teaches us something. It won't be long until we have both the technology and the knowledge to actually track them down and arrest them.
And we've had some success in that arena, too.
I think the main thing here, is this:
You have everything to lose by attacking a company on the internet. The bigger the company you attack, the bigger the thing you are risking.
A large company has NOTHING to lose by tracking you down. Sure, it might cost it money, but they have plenty of that.
You might think it's a great laugh right now, but when you're arrested and taken to court, and suddenly a lot more is on the line than your reputation amongst the other kiddies on IRC, I think maybe then you will regret even getting involved.
It's not cool, it's not elite, and we will catch you.
Only problem with that is if you block DDoS at your router, it's still wasting your bandwidth! That's enough for most DDoS.. Our link (a T1) couldn't overload any of our web, dns, or mail servers (Solaris, Linux, FreeBSD, OpenBSD, NT, 2000) in terms of CPU, net or disk I/O, but doing that would significantly fuck our users trying to do surfing, mail transfer, etc since their traffic would be lost in the crap.
:(
The only fix for that is to have your ISP's end of the link put in rules, which depending on the skill of the DDoS'er or his scripts, would block out most incoming traffic, including legitimate inbound requests from clients, telecommuters, etc. The real sneaky thing about smart DDoSers is that they forge IP addresses from all over the IPv4 space and so you can't actually tell genuinely which net they're sourcing from without an extensive & laborious backtrace or software that can perform such backtraces by negotiating between peers automatically..
Pointy-hair summary: It's ugly any way you slice it
Your Working Boy,
Why don't sysadmins start blocking off invalid TCP/UDP packets that the router?
Because many ISPs can't be bothered. UUNet, for example, refused to do backtraces on a DoS attack on my network at all. Multiple ISPs would need to cooperate in a very thorough way, and they don't see the $$$ in it I suppose.
I know at least 1 piece of software (Manhunt) is looking to get installed within ISPs to monitor routers and automate this backtracking and concomitant inter-ISP coordination, but I don't know if that's gotten anywhere.
Your Working Boy,
Spike strips usually consist of hollow spikes that break off and stay in the tire; "self-sealing" tires are no defense. There exists a superball type rubber compound you can use to fill tires instead of air; it is commonly used in construction equipment and tractors but will fuck up the handling and ride of cars on a highway.
Interestingly enough though, you can blame Slashdot for inciting DDoS attacks. When the editors post articles claiming that such and such company did something bad, you will often see comments (highly rated!) saying "let's DoS them" and even posting scripts to do it. I didn't take this seriously until one day Slashdot decided to pick on a place where I worked and suddenly hundreds of DoS attacks started.
This kind of thing doesn't exactly help with the hacker/cracker distinction that Slashdotters seem so keen to enforce.
But the fact of the matter is that ORS 164.377 is overbroad and vague, and that the police and judge created a search warrant out of speculation. We are arguing that in court right now, and the jury is still out. Until that matter is resolved, the fat lady hasn't sang yet.
And in the meanwhile, Oregonians (and residents of many other states with very similar overbroad and vague laws) are at risk, for doing their job. I've had dozens of people come up to me and say "there but for the grace of God go I" over the past seven years.
Yes, I did stupid things, even with good intentions. Perhaps I should have gotten fired or worse. But being made a triple felon (and losing a cumulative year of work and a quarter million dollars) in the framework of bad legislation and bad implementation doesn't fly, and I won't bow down to it.
For more details on my ongoing case, visit the FORS archive.
Acutally, I prefer Spider Robinson's analogy that this DOS attack was like "a 12-year-old nincompoop gluing shut all the doors of the mall". Very appropriate as most of the sites are nothing more than commerce. I don't much care if I can't get into a store.
I forgot to add a link to the full Globe and Mail story.
What should really happen is that ISPs should demand that egress filtering take place at the client side and charge extra for connectivity -- a DoS tax, if you will -- to sites that will not egress filter or refuse to provide proof that they ARE egress filtering.
Even with this type of protection, there may still be topologies (DSL? Cablemodem?) where egress filtering may not be either possible or practical until much higher in the food chain. If the ISP in question is a large one, successful spoofing may only require spoofing that ISP's CIDR blocks or other addresses that could pass an 'exit router' egress filter.
I'm not sure that there is a real solution to DoS attacks with many current protocols. Requiring a brief client handshake/auth mechanism may be the only solution, and that makes the net a whole lot less anonymous.
If he knew the details of such an attack, he would also know the severity. How does a "not very technical" person outline a DDoS attack. You have to have a decent understanding of TCP/IP.
And the behavior of ICMP protocols. I cant imagine someone who has been using the internet for a few years to surf and email understanding what mixter wrote in his papers. What normal users do you know of bought a book on TCP/IP and even know what ICMP stands for? No one I know.
Microsoft aggravates my tourettes syndrome.
We actually tested this here with a default install and a basic firewall setup. What we thought a user who just wanted to be a user and not anything more would do. It worked quite well, the host had all ports filtered. incomming UDP/TCP. ICMP still worked.
Microsoft aggravates my tourettes syndrome.
You obviously have no clue how Microsoft's .NET architectures work. Data is stored locally, as well as on the server. The whole point of having the server in the architecture at all is for replication to other machines and/or devices. So, no, you won't be writing your monthly report online, but when you save it, it will get replicated to the server (and probably at intervals before you save it, as well, as a work-in-progress).
So, some skript kiddie takes out the connection to your .NET server. Maybe you can't get your e-mail, but any documents that have been replicated to your local store, and any documents you're currently working on, will be perfectly accessible. Any changes you make won't get replicated to the server yet, as you can't reach it, but at that point it's back to pre-.NET business-as-usual. You'll just have to do "old-fashioned" replication to your mobile devices and such.
There are two companies that look like they are approaching this space with interesting solutions that have a high probability of working. V-Secure and Mazu Networks. Interesting players to keep an eye on.
Really? I was sure I knew exactly what one was WAY before the 7th of January last year... But maybe that was just me. In related news: RedHat's new Beta release has something that I've been bitching about for years: a NETWORK SECURE client install!!! At last, someone who doesn't know what they're doing won't have RPC, statd, named, telnet, ftpd, and all the others running on their machine for some 1337 moron to crack into to use as a dDOS tool. Took them long enough.
hmmmm?
Closing the implementation would do nothing to enhance security. It just makes analyzing and fixing problems and preventing attacks that much harder.
The things that would actually make a more secure protocol - controls to prevent spoofing and protocol-layer encryption - are in no way easier to accomplish with closing the implementation. In fact, they will be the better for the openness.
And, it's worth pointing out that the openness of TCP/IP has allowed it to become the de-facto networking protocol, period, beating out closed candidates that were arguably better protocols.
F Jackie.
Boss of nothin. Big deal.
Son, go get daddy's hard plastic eyes.
Expanding a vast wasteland since 1996.
Perhaps today would have been a much better day for those UBC students to do their Golden Gate prank. Right place: San Francisco, heart of the dot-com craze. Coulda used Cat-5 cabling to hang the bug off the bridge :)
Does it make you happy you're so strange?
You have my thanks for informing me of such a cool Spider Robinson quote that I was unaware of. *tosses a mug into the fireplace in your honor!*
Cheers!
--
Remove the rocks to send email
On the whole, I find that I prefer Slashdot posts to twitter ones because I don't get limited to 140 chars before
I'm thinking that a properly oriented high power snowblower-like device, mounted in front of your tires, could probably do pretty good job against at least small-time Tackers. Those who put more time and effort into their work (by sticking nails into 1x8 boards or pounding their tacks into the soft pavement) would be harder to fight- but it could be done. Self sealing tires would be the obvious easiest solution, although you could have lots of fun combining directed energy/acoustic weapons, liquid nitrogen, and road analyzing radar.
The roads must roll!
... and there is no doubt, that one day he will be
where the eye of his telescope has already been
take a gander:
http://www.mazunetworks.com/
Marketing info states:Mazu's technology is uniquely suited to solving the DDoS problem because it enables a proactive, focused and intelligent approach instead of the after-the-fact, fragmented and manual methods that most businesses try to employ today. With Mazu, businesses can outwit, outflank and outplay DDoS because, for the first time, they can operate with more fine-grained knowledge and resources than the attackers.
Well now, that depends on the velocity of the thumbtack, doesn't it? (Hmm... I think I have a new project this weekend.)
First Saturday in April is the hash bash in Ann Arbor MI, so that wont work.
IMO, unfortunately the best prevention method for prevention of DOS attacks requires work by people who generally don't get attacked. Yahoo can't truely protect itself, it is the hundreds of insecure server operators that must work. Perhaps ISP's should work with server operators to make their servers better equipped to prevent an entry by a nefarious source...
-Moondog
YES! I was entirely serious. As another person commented, there are already unrelated organizations and taskforces, but we need one single international Internet police force!
I think the solution we need is some sort of "CyberCop" organization. This group could be involved in tracking down online criminals who engage in DoS attacks, web defacing, and other terroristic acts. The FBI is obviously not doing their job, so someone new that is completely dedicated to Internet crime could help.
Additionally, this organization could set guidelines for ISPs, like requiring them to keep tracking information on certain packets for a period of time, or requiring them to block packets from unrouteable addresses. @Home is horrible about this. I've noticed routers in the 10.x.x.x subnet upstream from me on the @Home network! That is unacceptable. What happens if that router tries to send information to my computer? It gets blocked by my firewall because I don't allow IP spoofing!
Anyway, we NEED CyberCops to enforce laws on the Internet. Maybe we can get other countries besides America to help pay for it too. That could give them some say in the rules were.
I'm not going that far; what I'm saying is that if I'm running an ISP and I know that I own the 23.45.67.* block, I won't let packets with a source address of 98.x.x.x or what have you out of my network.
I'm not saying you need to validate every packet that comes out (way too computationally expensive, i imagine), just that the same way you set up ingress filters preventing packets with a return address of 127.0.0.1 or 10.x.x.x and whatnot come in, you should prevent those obviously falsified packets from ever going out.
Easy does it!
This comment has been submitted already, 276865 hours , 59 minutes ago. No need to try again.
- ifconfig wouldn't show promiscuous mode,
- netstat wouldn't honor -p and
- ls, ps, find and du were also modified.
Amazing.Moral: the script kiddies are totally indiscriminate. Once you're connected, you're vulnerable if you haven't taken protective measures, which include applying patches or upgrading vulnerable software, turning off un-needed services and firewalling/packet filtering. File integrity checking is also a good idea to warn you ASAP of a compromise.
For nomination(s) to 'assholes of the century', how about the schmucks who write the scripts that script-kiddies use? Why would anyone do this?
"that's not encryption - it's a new perl script that I'm working on..." - from some Matrix parody
"I mean, who on Slashdot was really freaked out when the Yahoo DDoS happened?"
Just because it was Yahoo, does that makes it ok. What if it was your online brokerage company that was being DOS'ed and you couldn't get through to tell you're broker to sell your RedHat stocks before they evaporated?
Does your above statement still work?
Hold on a second... Steve Bellovin's area of research consists of DOS attacks. Should we immediatly suspect someone who does research in this area. This is exactly like saying all Sys Admins are script kiddies. If he were such a suspect I can assure you that he probably would not be chairing the ICMP Traceback working group at the IETF. Come on Slashdot... this should practically be flamebait. Let's not chastize the researchers that are trying to prevent DOS attacks.
A. Keiper
i have seen many suggestions, such as online cops and such, but i got a good idea that might work and might get some sys admins mad and actually to take responsbitity for their comprimised boxes.
well here is the idea i came up with if there's a known script kiddie or comprised box the admin refuses to deal with that network should be blackholed, don't allow them to route to any place outside their own network, until they can prove it has been fixed. some admins won't listen to another person screaming at them to fix THEIR problem, i know this from personal experience, but give them several hundred or thousand paying customers and people who pay them yelling to fix the problem and then you will see how fast things change.
Becouse the term does describe an individual that has a strong disire to learn how things work no matter what the context, and this is being turned into a negitive attribute of the citizen. Thus encourging the citizen to give up his rights to figure out how things work. This ofcourse keeps the money at the top where the people who hold the knowladge reside.
Hey where I grew up a hacker was someone how didn't RTFM. In other words it was someone who taught himself. You want to know how that car works well go out there and take it apart (Unfortunitly this will soon be illegal)Its not about cracking or whatever.. it's shouldn't even just refer to computers.. hacking is about thinking and about having a unusually strong disire to learn. To understand the universe and not take things for granted.
YOU try filtering 100+ MB/sec of traffic and tell me how well YOUR router handles it. Make sure you write about a hundred different rules which are applied to every incoming packet.
--
ALL YOUR KARMA ARE BELONG TO US
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
Mess with the internet on the other hand and you're a force to be reckoned with.
You only become a credible threat when people believe you can hurt them again and again and again, whenever you want to. That's what it takes to be "a force to be reckoned with."
Even assuming that you aren't arrested shortly after taking down the root servers, you have to be able to convince everyone that you can and will cause similar havoc again and again.
But all of these holes are one-offs. Every time you abuse one, it will be fixed. You would have to convince us that you can invent new exploits faster than we can fix them.
...i'm using "you" figuratively here.
Was it a forgettable stunt? No, It was funny event. A much-needed wake-up call for insecure e -commerce sites? Yeah If a young kid with very little skill could do that, to a big huge company. I magine what a really skilled vetran of the scene could do if that person really got pissed off. Lame script kiddies giving hackers a bad name?The only things those attacks was to KEEP THE FOCUS on the Geek Community. Much un needed focus. Things are hard enough. We got things like -- "The mind boggles. Police have apparently raided a student's dorm room due to his participation in a heavy metal music inspired Starcraft clan, 'Bled For Days.'" Posted by jamie on Tuesday February 06, @03:15PM from the darkened-hearts dept. -- Just because a person thought a file was a threat. Not thatstudent meant any harm. There a real fear out there when it comes to computers(thatnks to the media) and its only gonna get worst if Lame script kiddies keep F*$#ing around....
If we refuse to be flexible, we are in effect opting out of the game of life. The world moves on without us.
stupid firstposters
Ever wonder what real hackers such as Theo de Raadt (OpenBSD), and Alan Cox (Linux kernel), feel about this?
They are, after all, real hackers ...
CNN tried to get it right back in 1999 when they interviewed Emmanuel Goldstein of 2600, but then they interviewed this guy from IBM and forever got it wrong...
:/
Anyway, I had written up a whole history of the term 'hacking' on CNN, but then Netscape crashed and I am Not a Hacker so I can't really retrieve it all that easily. I WAS a Hacker, but that was Fortran on the DEC... *sigh* I can't keep up with hacking anymore... which may not be a bad thing if hacking is so evil...
Here's the links:
The Palmer Guy
Goldstein
SDMI: Finally! Music that won't rip or burn! Brought to you by the fine folks at RIAA.
Lame script kiddies giving hackers a bad name?
/. readers that we respect someone who does something that is hard and takes a lot of effort and creativity.
Hmm. It is the general belief of most
From that mindset, the person (or people) who first thought up a DDos attack are to be respected, since you must admit, it took some skill, programming, hacking, and theory to get it accomplished. (Note I leave morals out of this)
Yet who is REALLY to blame when a Script Kiddie does a bad deed? Personally, I blame the idiot who MADE the toolset easy enough for a mindless goon to use.
------
Let me give you the lowdown
I thought there was some protection against these attacks included in the new linux kernel networkings.
Liberty.
methinks i have not yet been caught, all those lame dot-com startups have been taught, lame_ipo.com has met it's end, and buy_pet_puke.com will have to fend, for customers who will shell out their money, to buy a ton and a half canister of honey, alas the tale must come to an end, over my shoulder looms a teacher
Really though, DoS (or DDoS) attacks don't do anything except spank the owners of the site for not protecting themselves as best as possible, no? It's expensive for them, yes, and nothing's perfect, but as far as I know, it doesn't cause other vulnerabilities; so it seems to be a matter of convenience for most sites.
Perhaps I'm just insanely naive?
With years and wisdom, fifty winters
A king, when a dragon awoke from its darkness...(92)
He's dead, Jim. You grab his wallet, I'll grab his tri-corder.
Haw. It'd solve the wave of denial of service problems for sure.
too bad you can't spell frontier.
on a more related note, 'online cops' won't do anything. have the police helped much in the physical world? as the number of enforcers increases, i'm willing to bet that the number of reactionary, psuedo-rebellious, angst-ridden script kiddies will increase too.
i don't let the man tell me where to sit on the freakin' bus, so i'm not gonna let him tell me how to compute! (or something, and some stuff. . .)
!-- wit --!
o, how about adapting some of those cool punishments in tartarus, in the Aenied, or how about stealing from Dante's inferno?
We force a script kiddie to sit in front of a computer with a can of jolt and a box of twinkies. Whenere he reaches for the twinkies, they move further away, but the computer moves closer, whenever ehe reeaches for the coputer, the twinkies and jolt move out of reach! He can never have the twinkies, the jolt or the computer, they're all just out of reach!
how about forcing script kiddies to run vigorously with several hundred pounds of antiquitated computers strapped to their bodies? now, to add some fun to the whole thing, we could loose some rabid dogs!
i think my favourite punishment from the inferno was for heretics, maybe this applies to some script kiddies too. divergent computing practices, divergent religious beliefs, it's all the same, right? We could shove them in a hole, upside down, with their feet sticking out. Now, we can't light them on fire, because that would kill them, but we could fill the hole with something that itches (maybe wool), and they can't scratch the itch, because the hole is too narrow!
just a few ideas. i'll leave the rats and racks for another day.
!-- wit --!
150 mph isn't that fast.... depending on the skill of the driver and the type of road. Don't belive me, look at nascar or drag races. My motorcycle takes me way past 150 mph but I've never killed anyone because I wasn't in an area to do damage. I was on a track that was equipped to handle such speeds. Now for the segway to the net.
The internet is like a race track if you can't handle the load you need to, then get out of the race. Slashdotters are legitimate people (except for the trolls) that wish to view the page, if the site can't handle it perhaps they need to reevaluate their site. Slash dot irresponsible I think not, more like irresponsible web hosting.
Are you lonely? Hate having to make decisons? Meetings, the practical alternitive to work.
Lame script kiddies. All they had to do was download a DDoS proggy, then upload to many choice workstations (probably a school's computer labs). That wasn't hacking. Now DeCSS, THAT's hacking!
"Ancillary does not mean you get to rule the world." --U.S. Circuit Judge Harry Edwards, speaking to the FCC's lawyer
It's also other media as well, television, the movie industry. Journalists who don't have a clue should not write tech stories. Since more likely than not they try to relate information with either something totally different, or inaccurate terms which they don't understand. I think as a community, any group that's gotten bad press due to stupidity should stand up and demand an apology or at least start a petition that the journalists be better educated. Symbolism can be used for good as well as bad, people need to take back symbols that had their real meaning perverted.
DNS wasn't designed with true builtin redundancy in mind, which was the whole point of DARPA. Freenet attempts to add redundancy in addition to privacy.
As Ian Clarke of the Freenet project has said, the whole DNS system is a poor hack that reduces the strength of the internet. Using a distributed system like Freenet will make DDOS inconsequential.
...have a nice day. =)
/.
Either that, or this is highly sarcastic and not very clearly written as such. However, a sarcastic piece disguised as a serious one is ALSO called a troll, because it's designed to catch the unobservant and hasty posters.
"...throw cash at the problem"
No one uses language like that except to argue AGAINST something. Not to mention the reference to the Tower of Babel, which seems like a sly joke to me.
"The internet needn't be a lawless frontier anymore"
Anyone who posts as much as this guy does has GOT to know what effect that sentence will have on
Thus, the above post is: A troll, a flamebait, or a moron who after posting a TON still can't see that this sort of thing is inflammatory here.
So what's with "5, Insightful"?! Maybe "4, Nicely Subtle Troll".
-Kasreyn
Kasreyn: Cheerfully playing the part of Devil's Advocate to hairtrigger
The only defense against DDoS attacks(and DoS attacks in general) is to distribute important servers and services over different networks with different IPs and ISPs. Keeping your webservers all on the same network segment is suicide for a big company that needs uptime. Same goes for any other services like DNS. without a central target to attack it becomes much harder to take down site. Use numerous isps and mirror the website(or DNS records) on seperate servers, it really isnt that hard to do and the costs are minor at best, definately worth the investment for companies that need 100% uptime. Couple this with a good routing setup and competent sys-admins and you have the best defense against any DoS attack.
Oh Well, Whatever, Nevermind...
See my article about this...
Mark my words...if anybody besides geeks read this site, by next week you'd see somebody who'd actually done this...then picked up on this article and referred to it in the newspapers.
"educate the public"
About computers, you mean? THAT'S your first hurdle...once you jump it, let everyone else know.
Not if Ketch has his way. The problem with this solution is that lots of people would LIKE that punishment. We'd see a huge influx of masochists in IT land, cracking just badly enough to insure they got caught.
Then they'll hire sadists to track and punish them. Oh, wait, they already did.
There is a thin line between genius and insanity. I have erased that line. -- Oscar Levant
But I'm not.
So I won't.
The outcome of DDos / hacker attacks, will only lead to better security. Look at the advancements in personal pc security and how many more people are aware of them ( black ice, Zone alarm ...). Also take into account the affects of mafiaboy's action related to CERT's recent annoucing that they will broadcast the exploits after a certain amount of days. ( i think it was CERT ).
The greater good of script kiddies and hackers is that they make people aware of the weakness. And knowing where your weak is the greatest advantage you can have to secure your system beter
spambait e-mail
my web site artistcorner.tv hip-hop music news
please help me make it better
if you see me, smile and say hello.
uhh, the comment was meant to be a joke. duh.
CmdrTaco is an idiot.
Through the perception of illusion, we experience reality.
(Sorry for replying to an invisible article; I'll slash my dot this afternoon as penance; moderate me 0 for this or I won't respect you in the morning).
And we return to my point.
Randal should not even have been fired for these things he did. He was employed by Intel at a time when network hacker cowboys did things like maintentance backdoors and idle crypto noodling as a part of their job, which they were basically making up as they went along. Then Randal pissed off the wrong guys and those guys realized they could throw the book at Randal.
And Intel didn't just succumb to the intervention of the police. They encouraged it. They inflated the estimated cost of "repairing" the systems Randal was accused of debilitating. You know the drill. Haxx0r X pings Yahoo.Com and Yahoo.Com claims it cost them $Umpty million in downtime and scour to find the heat signature of the ICMP packet.
When your doctor removes your appendix as a prophylaxis during a laparotomy, do you charge him with assault? If the police intervene, do you encourage them?
So, again, metaphors intact: Intel put Randal Schwartz in jail for doing his job.
--Blair
Intel chased Randal down the hall because of back-doors he'd installed in the Intel network (to make it easier to fix the many broken things when he was physically off-site). His diddling with DES/whatever cracks on ora.com's password files was clearly recreational at a time when anyone who knew how crypt() worked had done it (including me, and I'm as anti-wrongful-entry as any computer user can get; but playing with numbers and benchmarking entry holes doesn't equate to B&E any more than whittling equates to mayhem).
All the Intel brass had to do was tell the Oregon cops that it wasn't a big deal. They didn't. They supported the polizei's efforts to cruel and unusual all over Randal. The rest is ugly, brainless history.
--Blair
Checkpoint software (Nasdaq's CHKP) claims to have some sort of defense against that sort of attacks... are they wrong, or did I hallucinate? -Alphaparadigm
-=The Dude=-
What if someone decided to fire missles at us while at the same time flooding all our military computers with packets such that we couldn't respond?!
Not that I'm in favor of a nuclear holocaust, mind you, but I figure the military brass would be particularly interested if their massive computers could get knocked off the 'net by a couple teenagers.
Tacks? Silly boy. Didn't you see Blues Brothers 2000, it's roofing nails.
What is pirate software? Software for inventory of stolen treasure?
I don't really see a way to prevent DoS or DDoS attacks in the future. The concept is way too simple. All you have to do is flood a site with enough packets to overload the server, and you've committed a DoS attack. The DoS attack will probably continue to evolve into something that isn't as esily stopped as some people mention. The DDoS will probably never be stopped because no matter how alert you are with security patches, there is still the likelihood that a hole will be found and exploited before it is patched. Once the machine is cracked, that's it. With the number of machines connected to the net, you could easily gain access to enough to start a DDoS attack once you find just one such security hole. You can't punish them because companies don't have jusidiction in foreign countries and as we've noted several times, each country appears to look at things differently. IMO, the best you can do is use the most secure OS you can find and pray. Am I missing anything?
Khyron
This is something that Slashdot does. When it posts a link to a small site, it knows, from experience, that this is likely to break the site.
This is no different from a script kiddie attacking a site. Both are acting irresponsibility.
Slashdot knows that by posting the link it is likely to cause serious loss of function - so do the hackers. It's like driving your car at 150mph and then saying "You can't blame me because people are dead". It's a silly argument.
Slashdot is being hypocritical by posting this story, considering it hass killed more sites than any script kiddie.
Excuses are not good enough. Slashdot has no obligation to link to sites. Other business are forced to act in a responsible manner - they don't let factories release any amount of chemical they like, and so it should be with Slashdot.
I really don't like this hypocrisy.
If you're that worried about efficiency you should use Morse Code.
A document with two spaces in it is probably 19,460 bytes in size. One with twelve words is 19,600 bytes. There's a constant 'header' minimum for any Word document. That seems reasonable. There's a miniumum size for a LaTeX file with one space in it, too. What's your problem, then?
I can't believe no one has taken down the root servers yet.
The attorney general went apeshit just because of Yahoo.com and e-trade. Imagine what would happen if the *.root-servers.net suddenly stopped responding. 99.9% of internet users would be paralyzed and helpless.
Here, instead of releasing poison gas into the subways or toppling the world trade center, this is really easy to do and americans will so get their panties in a bunch:
- Amass lots of rooted boxen (given). Use the BIND exploit for the ultimate irony.
- Write a perl script and use a resolver module to send bogus random requests to each root server in sequence. The more random the better, as they will be harder to filter. Don't forget to spoof the source address.
- Run on each rooted box in background. Cron it to start on boot.
- Gloat to world newspapers.
It's ludicrous that none of you extremist terrorists have done this yet. You can do this from the comfort of your own homes and you don't even have to risk capture if you live in a US hating country.Killing a bus full of passengers is good for horrifying headlines, but in the end no government will really care. Mess with the internet on the other hand and you're a force to be reckoned with.
And for all you jackasses crying Treason, would you rather they poisoned your local water supply or that they just took down .com? I know what my priorities are.
The DDoS attacks last year relied on the ability for Mafiaboy to install programs that would help propigate the DDoS across a large number of unintental volunteers' computers, such that all he had to do was wake them up at a given time with a given target, and that's all he needed. He was able to get such programs installed thanks to the help of email viruses, web page javascripting, and activeX. IIRC, many of the computers that were found to be part of the attack were computer clusters at universities, implying how easy it was to get this propigated.
If we had OSes and browsers that would not run untrusted code unless the user said yes, the DDOS would not have had been as effective. Even if that option's there, the important of what untrusted code is is not well implied. MS's 'error' message if you use prompting for ActiveX controls and scripting is "Scripts and ActiveX controls are usually safe..."; this is NOT true. Sandbox the browser, do not let it access any system files (as there's need for it to!). And make sure that computer users KNOW this and the effects that running such programs can have, don't take a passive view of "oh, a new bug fix is out, you ought to install it when you get a chance...".
"Pinky, you've left the lens cap of your mind on again." - P&TB
"I can see my house from here!" - ST:
What the article doesn't mention is that is father is called up on the witness bench, and his name has come up in the list of acusee, as he is, according to the procecutor, probably involved in the DDoS attack.
MafiaBoy's father allegidly gave him information on the technicalities of such an attack.
Local newspapers have reported at some point during the year that this is what's going to be used as a defense. The father allegedly knew how to do such an attack, for having read about it, and discussed it to his son, which then tried it. The father did not know the extent of the attack, not being very technical himself, hence the defense relying on the fact that MafiaBoy did not know either that this would cause such a severe attack.
Another newspaper had reported that the kid itself was "frame through ignorance" by his friends to do th3e attack itself.
Both newspapers were full of inaccuracies, of course, such as for the usage of the word "hacker", as usual.
Karma karma karma karma karmeleon: it comes and goes, it comes and goes.
Once people start combining attacks with stock market manipulations, people might start paying more attention.
Sell short EBay, DDoS them for a couple days, collect some cash. Day trading and the speed at which attack news travels has made the markets so much more reactive to the slightest bit of bad news. Do this just before some kind of major EBay event so you can claim a legit excuse for the sell and hide your tracks carefully when starting the DDoS (AOL via a stolen cell phone?)
You heard it here first.
c.
Log in or piss off.
If you did that, would that make you a Tacker?
Then the media could go bonkers about attacks by crazed teenage Tackers out to bring down the highway system!
(But seriously, a thumbtack wouldn't do sh!t against a car tire...)
Does it make you happy you're so strange?
Well, it depends on your tire design I suppose. If you compartmentalized the tire well enough and used some kind of emergency reinflation system like that "great stuff" expanding foam to refill the punctured compartments...
... and there is no doubt, that one day he will be
where the eye of his telescope has already been
Oh, yeah. It all went back to real life, where this is no more than some offended 5kr1p7 k1dD13Z deciding to lash out. It had no influence on the world as a whole, had (as the article pointed out) no influence over the cyber-world...
This was an event that didn't shape anything. It didn't cause any sweeping changes (i.e., Columbine or the Challenger explosion), and certainly didn't bother anybody a week after it happened. I recall being astonished at the organization, having so many people DoS-ing at the same time... it gave me hope that the Internet community could bind together and fight for a common cause. Instead, it was just a trojan run by a single person.
It was a non-event of Y2K proportions. Get over it.
------
Like sane egress routing checks set up on the individual ISPs end?
No, it won't prevent DDoS attacks, but if the checks are set up so as to prevent packets with spoofed IPs from ever leaving their segment, then the people being attacked can see who's attacking, drop packets from them and notify the ISP hosting the (inadvertent?) attacker, letting them know what's happening.
Easy does it!
This comment has been submitted already, 276865 hours , 59 minutes ago. No need to try again.
One of the latest developments in the war again DOS attacks has been with a working group at the IETF that is trying to create ICMP Traceback messages.
Essentially what these messages do is generate an ICMP packet with the previous IP address and the present IP address with, I believe, the first 60 bytes of the packet for every 20,000 packets that pass through the router. This packet will be sent to the source address so whoever the poor victim is can figure out who the REAL culprit is and not have to chase after spoofed IP addresses. Of course this should only be done on the edge routers and not the core so as to not generate unnessary traffic and to keep the internals of a service provider secret.
Now when this would happen is somewhat up in the air. Those of you that have attended IETF meetings know how slowly things can move (my personal experience is with diffserv... shudder, 4 years to argue about 6 bits of data in the IP header). Not to mention every single router vendor has to implement this and on top of this, the service providers have to update their routers with the software updates that support ICMP traceback messages.
May Day might be a historically consistent day for rebellion/mischief/etc. Hey, it works for the anarchists and whatnot, no?
Problem is that these "internet trash" have exactly 0 respect for rules to begin with, so thinking that all of them (or probably even a significant portion of them) would abide by the one-fun-day-a-year approach is probably optimistic. Cool idea though! :-)
--
Fuck Censorship.
News for Geeks in Austin, TX
We all know these sites weren't DOS'ed. They were Slashdotted!
Men believe what they want. - Caesar
While the state of the art in withstanding an attack has advanced measurably with the new kernel (SYN cookies, etc.), the Ramen Worm and other recent security problems have shown pretty conclusively that it takes a long time for security patches and package updates to make it into production servers.
Unfortunately my friend this has nothing to do with OS kernels, and everything to do with infrastructure elements like pipes, routers, switches, and firewalls.
The infrastructure cannnot handle the level of load being placed on it when these attacks take place.
I agree you can actually DOS a server, but these attacks were against the infrastructure.
Why not have a DDOS reunion tour? I'm sure the folks at CNN, Ebay, etc would love to see your sup3r 1337 skillz again...
morons...
I am become Troll, destroyer of threads
The rush-to-market took presidence over security, even though preventative measures against DDos attacks was outside the remit of most sites, it was a wake up call.
A year later secuity is a lot higher in the product requirements!
Okay. So, it's basically DNS that ships around Word documents instead of zone records...
Hmmm... Opening Word, hitting the space bar once, and then saving the document creates a file that is 19,456 bytes in size. (Under Word 97, Windows 95B, using the normal.dot template.) Adding a few generations of Microsoft Bloat, multiplying it by millions of proles... afraid to estimate the implications of PowerPoint...
Sounds like, through sheer volume, it might create its own DoS attacks...
;)
Fire and Meat. Yummy.
There's a miniumum size for a LaTeX file with one space in it, too. What's your problem, then?
Bloat.
with two spaces in it is probably 19,460 bytes
(2/19,460)*100 = 0.01027749229188% efficiency.
Hmmm... I think that's even less than I expected from a Microsoft product.
And when files like that are being passed around between .NET machines the way zone records are for today's DNS servers, I worry about the future Internet traffic.
I think I'll stick with vi for all my text editing needs.
Fire and Meat. Yummy.
How are we to protect ourselves, and save the new economy and way of life and working we see growing for the first time?
Yeah! But if Microsoft moves all of, for example, Office 2003 to their ".NET" philosophy before DDoS has been conclusively thwarted, they're shooting themselves in the foot.
Who is going to buy into .NET when any 15-year-old with a cable modem can lock every secretary in the world out of Word? Every accountant out of Excel? Every CEO out of PowerPoint?
(Okay, not *ALL* of them, but it will be enough that almost all global business stops at the mercy of a mouseclick over a WWF desktop in a New Jersey bedroom.)
The ease of committing a DDoS is therefore, in my view, a very convincing deterrent to the mass adoption of centralized pay-per-use software subscriptions.
Fire and Meat. Yummy.
Did anybody check this guy out? I mean, come on right?
The problem with capped Karma is it only goes down...
SIG: HUP
Reminds me of the allegory of the monkey who can't get his hand out of the cookie jar because he won't let go of the cookie. Security measures to help prevent all DoS attacks as well e-mail virus-like scripts and web scripts severely impair the ability to control and advertise. Until then, you're better off using 3rd party security measures.
----------------------
The sad thing is, e-business will probably decide that the better way to deal with events like these is NOT to secure their sites better, but instead prosecute the hell out of the offenders. That'll work well the moment someone else tries it and isn't too much of a stupid HaX0r to brag about it on a chat site. Also interesting is how these opportunities for learning generally end up involving the lawyers.
--------
Bleah! Heh heh heh... BLEAH BLEAH!!! Ha ha ha ha...
"Lame script kiddies giving hackers a bad name?" It's not the script kiddies giving hackers a bad name, it's the press's misunderstanding and misuse of the word.
http://www.codewolf.com - Just good stuff to waste time
Why don't sysadmins start blocking off invalid TCP/UDP packets that the router? AFFIK lots DoS attacks use packets with invalid TCP flags, have a look here. If they are dropped by the backbone provider end of problem..... mind you having said that most crackers will simply find another exploit.
I suppose some sort of stateful tracking would be handy as well, but that wouldn't stop DDos.
Its a game of chess
I can't think of anything witty right now
Its a dangerous attitude in some respects, but in others its not. Its dangerous because it makes folks think hacking is harmless (till their credit report gets ripped off, etc) But heck most people survive just fine if the power goes out for a bit, why not the Internet?
I'm not agreeing with them, I just see that in responses from folks I talk with that aren't /. readers. The scary part is, DDos attacks ARE the tip of the iceberg. Its kinda like a doofus with a gun. Someone fires one in the air, everyone runs for cover, life stops for a sec, and then folks go about their business, not caring if the bullet came down and killed some poor sap. It just leaves folks unprepared for the real deal like when hackers manage to cull sensitive info on many of the top public officials (or their comuter systems) and hold the government hostage. They'll be totally unprepared.
The best we can do is a) spread the word to our less technically inclined friends that it IS a big deal, b) hacking is different from cracking, and c) contribute to hack prevention/detection systems like Snort (Not necessarily in that order!
Top Most Bizarre/Disturbing Error Messages
I'm afraid the only way to make DDoS attacks infeasible is for victimized companies to begin suing both the owners of the networks that have been hacked to produce the floods of packets, and one or more of the ISPs responsible for forwarding those packets to the victims' networks. The grounds for such lawsuits would be negligence in not repairing security holes in those machines, and-or allowing communications from obviously spoofed packets inside their network.
When companies are informed of the potential liability of not properly securing their networks, they will finally take serious steps to prevent their property from being hijacked and used to attack other systems.
We all know this, but sometimes forget, so bear with me here. Hacking didn't used to have anything specific to do with security. Now it's all about security and how to circumvent it. Trying to call it "cracking" will never work. CNN has bigger disinformation pipes than the original hacker community, which has a "tiny urethra" of a PR pipe, and nobody wants to talk about that.
Mafiaboy is nonetheless the fall-guy for a worldwide Society Of Loners who will get the message just in time for their little sisters to find the crack pipe behind the auth server.
Meanwhile, national ISPs like WWC.Com and Frontier.Net can't keep their billion-dollar networks running for a week without a major outage. MSN hires gorillas who don't know Cisco from Crisco. Go.Com is its own worst enemy rather than the cyberjewel of the most widely held corporation on Earth. And Intel jailed Randal Schwartz for doing his job.
Cracking is relatively about as debilitating to the net as keying Vint Cerf's car. But I don't want to be associated with that, either.
--Blair
"My tan is the color of a television tuned to a dead channel."
While the state of the art in withstanding an attack has advanced measurably with the new kernel (SYN cookies, etc.), the Ramen Worm and other recent security problems have shown pretty conclusively that it takes a long time for security patches and package updates to make it into production servers.
Red Hat hopes to make a splash through their automated update services, but so far they don't seem to be making much of a splash.
What is really amazing is that there aren't more DDoS attacks, considering the continued vulnerability.
IMO, Urban Existentialists will be the curse of slashdot. They are becoming ever more frequent, and are frighteningly easy to implement. How are we to defend the moral upright citizens from attack when you can grab a hotmail address and troll away? Script Kiddies, with long winded trolls running amock, who needs 'em?
The e-economy is like a shining jewel, eh? Man, you smoke too much fuckin' pot, dude... lay off the weed.
My suggestion is to nuke your sorry ass off the planet, but that'd be unfair to those unfortunate enough to be near you.
News for turds, shit that splatters
The punishment for being a script kiddie who initiates a DDoS attack should be this:
Tie him to a table. then get about 20 people to stand in a circle around him. Then they should all converge on him, and poke him repeatedly. Just hard enough to hurt a little bit, but not too much. One person doing it would be annoying, but not bad. Multiply it by 20, or more and BWAHAHAHAHAHAH.
punishment for more serious attacks could replace sticks with finger poking. Lets see how long DDoS attacks would keep happening.
Of course, all of that would require that they actually put some effort into trying to find out who is responsible. All you have to do is get an infiltrator into some kiddie group. they like to bragincessently about their latest enterprise, whether it be leeching the latest warez release, or using 31337 sk1llz (some program made by someone who was actually semi intelligent) to h4x0r some computers.
For the people who actively try to crack systems, there should be a different punishment. If they get caught, they should be required to submit to a colonostomy. (To those non-medical geeks, a colonostomy makes a prostate exam look like a walk in the proverbial park.)Basically, they would be violated, and examined in the same way that they did to whatever system they got into.
Mostly script kiddies should obey my sig:
----------------------
Opportunities multiply as they are seized. --Sun-Tzu
Denial of service attacks are to cracking what parking a logging truck in the no parking zone in front of a bank is to bank robbery. It takes no talent, just a disregard for public convience and a big truck/pipe.
--
Remove the rocks to send email
On the whole, I find that I prefer Slashdot posts to twitter ones because I don't get limited to 140 chars before
I'm still wondering why the attack against Microsoft the day after they fixed their DNS routing mistake made so little news. There are still plenty of major web/e-commerce shops out there, but perhaps the spector of DDoS just can't make news and grab eyes like it did just a few months ago.
Evan - needs to hit preview before submitting
Outdoor digital photography, mostly in New Engl
Just because it was Yahoo, does that makes it ok.
No. It doesn't. In fact, Yahoo is my browser home page. I probably hit it dozens of times a day. As far as I'm concerned, it's the best all-around portal/search engine out there.
What if it was your online brokerage company that was being DOS'ed and you couldn't get through to tell you're broker to sell your RedHat stocks before they evaporated?[sigh]
Can we be rational about this for a moment? You write like you have exactly the same sort of momentum and hysteria going as NASDAQ in general did.
Okay. Brainflash: the Internet is merely a communication tool.
A DDoS interrupts your communication. Like walking into an elevator with a cellphone.
It's an outage, an interruption, inconvenient and frustrating but not the end of the world.
On the other hand, what would the ramifications be if someone could press a button and selectively give a cellphone user a brain tumor? (Oh, think of how useful that would be when you're driving!) For one thing, it would absolutely kill the cellphone. No one would use them.
This could be a parallel to more malicious and dangerous cyber-terrorism; breaking into secure machines and disseminating private information.
The DDoS is inconvenient and makes you reconsider your reliance on the medium. Hold the fire and brimstone: give your broker a call with a telephone.
Does your above statement still work?Unless the Internet is blown beyond all proportion, from being the (revolutionary) communications tool that it is to the realm of a lifestyle, yes, it does work.
A year ago, the Internet was basically down. The traffic from the DDoS was such that most other pages that I tried to load were unusably sluggish. At the time, I didn't know why. I pinged big sites (including Yahoo) and did traceroutes trying to figure out where the bottlenecks were. Satisfied that it wasn't on my LAN or even with my ISP, I gave up: Instead of looking up a supplier using www.four11.com, I picked up the Yellow Pages.
It sucked, it was inconvenient, I had dozens of users asking me why mail was bouncing and pages didn't load, but it wasn't the end of the world.
Fire and Meat. Yummy.
The defence is for the freaking administrators of all the main systems (major IP subnets) to not allow a subnet ping (a ping where every node that is alive returns a ping to the sender)...This would stop ALL DoS attacks in which the person causing the attack is only in control of one computer.
That leaves us with attacks that are comming from super-high bandwidth systems, and attacks that are using large numbers of systems. The high-bandwidth systems are MOST likely NOT going to be responsible for many attacks, as most hackers can not afford to pay for the kind of bandwidth needed. This leaves us back to the issue that the person starting the attack will need to break into any/all systems that start the attack. Now this could be easily resolved if people were just informed correctly about what security issues they need to worry about (like placing your system behind a decent firewall, software or hardware based).
That would then block out a very high number of the people trying to do these attacks, because face it, most of these attacks are from novice hackers who can not actually hack the system/entity that they have a problem with so they launch a DoS attack because it is so easy to do. Increasing the difficulty of launching this type of attack and the people who are doing these attacks will either need to learn how to be a better hacking (in which case they will probably find a way to actually gain access to the system that they are DoS'ing and just wipe them) or they will get fedup with it and go piss and moan to they friends.
How are we to protect ourselves, and save the new economy and way of life and working we see growing for the first time?
My suggestion is that we greatly improve punishments for script kiddies and throw cash at the problem by initiating 'online cops' with special dispensation to track them down. The Internet needn't be a lawless fronteir anymore.
Israel has done this to an extent. We should too.
You know exactly what to do-
Your kiss, your fingers on my thigh-
You know exactly what to do-
Your kiss, your fingers on my thigh-
I think of little else but you.
I wrote a text from the administrative standpoint on how to pretty much eliminate 80% or so of an attack on a variety of hardware/software based level which can be found at my site.
./script basis.
Now as for the attacks themselves, this wasn't anything new as DDoS became popular after Mixter coded a scriptkiddiot tool, which allowed malicious users to actually implement these attacks on a
The foundations for DDoS though are a bit old and could have long been resolved had thorough network's been set up to deny any malicious activity to leave their networks and attack others.
Many admins have the knowledge to do so, but I think theyre resources are tied into making things work right then and there as opposed to doing it right.
"When I was a Buddhist, it drove my parents and friends crazy, but when I am buddha, nobody is upset at all"
The linked article is out of date. On January 18th Mafiaboy pleaded guilty to 56 of the 66 charges. The other 10 charges were withdrawn. CBC has some details.
Regard these attacks for what they are irresponsible acts by people with little regard for the public good.
My other sig is extremely clever...