Slashdot Mirror
Linux Most Attacked Server?
Anonymous guy who can't remember his login sent in a story from the Globe And Mail that says "During August, 67 per cent of all successful and verifiable digital attacks against on-line servers targeted Linux, followed by Microsoft Windows at 23.2 per cent. A total of 12,892 Linux on-line servers running e-business and information sites were successfully breached in that month, followed by 4,626 Windows servers."
But think of how many more linux servers are out there than windows servers.......
On the surface, this statistic serves both as a testament to linux's growing popularity as a server OS and ammo for those windows admins who have long taken abuses about the insecure nature of their OS. These ideas, particularly the latter, however, may prove misguided; breaches against servers are rooted not only in the security of their running OS, but also in the effectiveness of the security implementation of the system admin him/herself.
It's ironic that Microsoft provides that service for free, whereas Linux requires paying money. But it's good because at least here there's a clear way to make money off Free Software and keep programmers like me from going hungry.
John.
More systems == more attacks. And consider that most servers are pre-configured with lame settings & passwords, and MOST newfi admins NEVER change the password... and if they do, it's a simple one that they use on ALL the systems they admin.
Geesh, it's not like we're talking broken protocols here...
The overall economic damage in August from overt and covert attacks as well as viruses and worms stood at an all-time high of $28.2-billion.
So while these "attacks" on servers totalling about the same damage amounts as usual there was quite a new record high obtained by the RPC vunerability...
So they are attacking an OS that is known to be running on more servers around the world and the "damage" from these attacks is holding steady, yet we don't mention in the article title that because Windows is MAJORLY vunerable, there was nearly 30 BILLION dollars in damage done!
Interesting spin.
During August, 67 per cent of all successful and verifiable digital attacks against on-line servers targeted Linux, followed by Microsoft Windows at 23.2 per cent.
/me turns off logging and closes eyes, going back to my happy place.
Of course, that really depends on how you 'verify' a breach, doesn't it?
*sigh*
During August, 67 per cent of all successful and verifiable digital attacks against on-line servers targeted Linux, followed by Microsoft "Windows at 23.2 per cent. A total of 12,892 Linux on-line servers running e-business and information sites were successfully breached in that month, followed by 4,626 Windows servers, according to the report.
Just 360 -- less than 2 per cent -- of BSD Unix servers were successfully breached in August."
I'm upset that they didn't mention the ratio of machines hacked... i.e. just because more linux machines that were hacked than microsoft doesn't mean that the ratio tells a different story. There might be more linux servers out there.
What is slashdot?
Also, it has gained something of a reputation as a secure system, at least compared to IIS, and this may be undeserved in installations where best security practices are not followed (most of them). This is perhaps a wakeup call that it's important to patch, only set up services that are necessary, and use a firewall and intrustion detection system, but most people know that already.
I never vote for anyone. I always vote against.
-- W.C. Fields
"Microsoft deserves credit for having reduced the proportion of successful on-line hacker attacks perpetrated against Windows servers."
The only way they've reduced the _proportion_ of attacks on their servers is by losing market share. The total number of attacks against Windows servers is still increasing, so it's a little premature to give them any compliments.
If voting changed anything, they'd make it illegal -- Jello Biafra
They claim a database of 280,000 attacks since 1995. They claim there were at least 18,000 attacks in August alone, or 6.5% of the total of 1% of their sample. Also, these numbers are meaningless without knowing the total population of each type of server. Oy!
I think it's time to break the statistics down application by application at that point. Show me some Apache vs. IIS numbers or MySQL vs. SQL Server numbers or exclude third party applications altogether please. For the record, I run both Windows and Linux for clients and servers and am pretty neutral in the whole OS wars thing. Each has their merits and uses, both need regular security maintenance and I am pretty much happy with both for very different reasons. I'm not a Linux zealot, but I know bad numbers when I smell them. And then...
So MS is shoring up third party applications then? They even go on to cite Sobig and MSBlast as the reasons for the high MS numbers. This is shifting over to a very FUD-like smell now.US Democracy:The best person for the job (among These pre-selected choices...)
I seem to recall some 500,000 servers being compromised by a worm last month. Do they only count attacks by people?
Well, that's sensible if you ignore the half million or so infections by Blaster - which clearly this article does.
I think that any analysis of digital attacks that filters out malware is missing a huge part of reality. Certainly you'd have to be nuts to call August a good month for Microsoft servers.
jim frost
jimf@frostbytes.com
I think a much more meaningful statistic would be how many fully patched Windows and Linux servers are successfully hacked. With Windows, you are always vulnerable, because the rate at which vulnerabilities are discovered far surpasses the rate at which patches are issued. With OSS, OTOH, a patch is usually issued a few hours or days after the vulnerability is discovered. Hence, the amount of time a successful Linux exploit is usuable is usually much lower than an exploit for Windows.
I would guess that most Linux machines that get hacked are due to unpatched/deliberately insecure configurations - like using a dictionary word for a root password.
The society for a thought-free internet welcomes you.
Number (or percentage) of successful attacks against servers maintained by professionals, sorted by operating system.
Of course there are a lot of non-secure Linux systems on the net. Lots of amateurs use Linux. After all, it's free! Notice how much the statistics in the article changed when they leveled the playing field and looked only at servers in one industry: government? Keeping to one industry caused them to look at systems maintained by sysadmins with much more equal skill levels.
From the article: Microsoft Windows servers belonging to governments, however, were the most attacked (51.4 per cent) followed by Linux (14.3 per cent) in August.
"We reject as false the choice between our safety and our ideals." --The American President (20.1.2009)
Windows hacked = windoze is ghey!
.conf files and convoluded iptables rules (4 of them to forward a port?!) that I can't really be certain.
Linux hacked = security is the responsibility of the admin!
Just look at all the backpedalling and but.. but... but..
Linux is not the super-secure platform you think it is. Not only because it's practically impossible to "not have holes in the code", but because it's a convoluded mess to try and configure.
Is my linux based router/gateway secure? I think so, but there are so many goddamn
I don't need no instructions to know how to rock!!!!
Folks who have traditionally been Microsoft users, who have recently installed Linux on an old machine at home or maybe as dual-boot, who have little to no real experience or training with Unix-like systems or with particular open source servers, are going into to the business IT environment and installing Linux-based systems on the hype.
Sure they can get Apache webserver serving pages, they can get Tomcat doing "something", and they can certainly run XMMS quite well on their workstation, but they really have no clue how to properly use these technologies in a production environment.
They see switching to Linux-based systems as being a simple fix.
They aren't willing to extensively review their configuration or product documentation. They aren't willing to put in the significant amount of time that is in fact required to become experts with the technologies.
Yes, they certainly do get a kick out of telling their friends that they have "Linux boxes running their shop", but security suffers due to their naive incompetence.
These techs should be fired.
Open source development may be a "we'll get that feature done when we feel like it" affair, but deploying Linux-based systems in a production environment must not be.
If anything, effectively and securely deploying Linux-based solutions requires more training and knowledge than does deploying Microsoft.
Let's stop pretending otherwise.
.sig Realistic fines for copyright in
We just had a bad infection of Nachi that hit about 500 of our Win2k computers, granted that doesn't add much to the already possibly skewed numbers, but it shows that they couldnt have counted every single successful attack against MS products.
Heck with just blaster and friends' numbers added to that, I'm sure that the linux number would be at least half of the MS number.
There were several hundred thousand computers compromised by blaster last month. Did they forget this statistic, or are they having steaks on uncle Billy tonight?
NM ;0
If you mod this as trollbait i'll kill you
Just an opinion and not trolling!
That's not the point.
The point is that this report handily debunks the myth that a Linux server is inherantly more secure than a Windows server.
The more rational among us here have tried to get the message out that no server is secure if there's an idiot at the helm.
Good admins make secure servers, not an operating system, despite what the zealots would have us believe.
"Ask not what your country can do for you." --John F. Kennedy
Anybody can into Windows, but it takes a real hacker to get into Linux.
Seriously, I suspect that difference comes into play when you look at where the servers are used. You'll find that Linux is used in more servers that are much more worthwhile targets (ie credit card transaction processing) than Windows. So going back to the original comment, not only is it less of a challenge to break into Windows, but I suspect that there is also less reason to want to attempt to break into Windows servers.
myke
Mimetics Inc. Twitter
statistics can be very misleading. for example:
Common sense can cloud statistical results. For instance, a technology firm discovered that 40% of all sick days were taken on a Friday or a Monday. They immediately clamped down on sick leave before they realised their mistake. Forty per cent represents two days out of a five day working week and therefore is a normal spread, rather than a reflection of swathes of feckless opportunists trying to extend their weekends.
(preceding was taken from an ars technica article)
if 90% of servers are linux servers, then it makes sense that 90% of attacks should be against linux servers, right? im pretty sure linux is more than 67% of servers right now, so 67% is actually very low!
RTFParent and learn a little bit about statistics.
If there were 100 successful attacks, 67 got through Linux servers, 23 got through Windows servers, and the other 10 are through other types (OSX et al).
However, this says nothing about how easily each system was attacked - we need to know the totals numbers, per system, of attempted attacks.
For instance, in my example - 100 successful attacks. Say there were 1000 attempts to hack Linux and 67 got through (93.3% secure), 25 attempts to hack Windows and 23 got through (8% secure) and 100 attempts to hack the others with 10 through (90% secure). Knowing just that the Linux boxes got 67% of the successful hacks tells us nothing. You need to know what the total attempts are to find how secure the boxen are.
-T
If over 12000 Servers were linux and were being sucessfully cracked compared to 4000 of windows boxes. Now representing this as 67% is to skew the results. What we dont actually know is how many were in the data set ?
Did they sample 20000 Servers ? 20,000 servers or 200,000 servers ?
Linux 67 Breached Linux Servers 12892 73.59%
Windows 23 Breached Windows Servers 4626 26.41%
90Total Cracked ? 17518
Well the percentile is only 90% of the figures. Which servers were in the missing 10%.
Did the survey compare windows to linux boxes alike e.g.
1 Linux Server examined to 1 windows box. for 20,000 boxes ?
I dont see any figures here for accuracy or qualification of the figures.
What I do see is a suggestion that Linux is very popular. If this is the case and we suggest that 80% of the net is unix to 20% microsoft. then 67% of 80% of the network being interupted seems very unusuall and rather high as a figure.
So I keep coming back to wondering where the figures have actually originated and been compiled.
Im fairly sure Microsoft can be secure, but unlike Unix it tends towards insecurity. Ive often compared running Microsoft boxes to herding sheep. You spend all your time keeping them alive and free of viruses. Unix on the other hand is the sheep dog, consistent , loyal and dependent.
They can bandy these figures all they like but unless they can flatten the survey and show a clear scope of investigation and comparison then I dont think we should be worrying about the quote.
And thats why Firecrackers and kittens don't mix.
During August, 67 per cent of all successful and verifiable digital attacks against on-line servers targeted Linux, followed by...
All the Windows boxes that are 0wnZ3r3d are not verifiable!
MjM
Groovy. Gear. Mod.
XKCD:Xeric Knowledge Comically Dispen
Almost all attacks can be solved by good system administration. All of the Blaster issues would have been mitigated if every windows machine was patched as soon as the patch was released.
Almost all software is insecure, if run poorly.
What is the life expectancy of a Redhat 7 default install not behind a firewall?
First these stats are important in demonstrating that it is important, no, imperative that admins of all flavors keep their servers up-to-date and know how to secure them effectively.
But, the stats also suggest that Linux is somehow less secure because it is attacked more often. The facts are a bit different though. Firstly, the statistics are drawn from a database of reported defacements not total defacements and definitely not total compromises. If this report were to be done in a more accurate fashion it ould have to include the hundreds of thousands of machines that are regualrly rooted by worms. Most recently, MS Blaster took over thousands of machines and reported for duty on an IRC bot channel. This report fails to account for these and many others like them.
I will conceed that Linux is defaced more than any other OS at this time but, I would also point out that this does not make it less secure. More people may report compromising a Linux box to change the Apache index page but, none of the Code, Red Blaster or many others bothered to register with the defacement database and I guarantee that these compromises outnumber Linux defacements by the millions.
I would say there is an important difference between server hacks and viri in that respect. Most people making a virus specifically target windows, while most people hacking a server don't target an OS, but an organization, therefore it is relevant that there are more Linux servers, while the number of MS boxes is not relavent in cases involving virus. The attack focus is different.
-3Suns
~~~~
The Revolution will be Slashdotted
This statement clearly states that less than 2 percent of the BSD servers on the net were attacked. Yet that is not what the numbers show. The numbers state that less than 2 percent of the attacks were against BSD servers. That is a very different thing indeed.
As such, there are a number of pieces of information that are needed to make this article useful:
The net will not be what we demand, but what we make it. Build it well.
It rhymes I think.
How strange it is that the numbers match up with the ones in this article.
I wish they'd split up the damage figures to clearly show the distribution. Just saying that 12,000 linux servers were bridged and 5000 windows servers totalling to 28 billions in damage including the damage done by the viruses kinda implies huge share of fault on linux. Worthless article.
Actually, I think it's not just "shared hosts" but also the fact all da pornz is hosted on linux. And even tho many sites are kept very secure, many more of those sites (numerically, probably the vast majority of them) can easily be "hacked" by something as simple as a referer spoof. And every one of those spoofed intrusions counts - ergo it's not just the lack of security, but the utter ubiquity of hacks that certain webmasters seem to want to remain exploitable. Pretty sad when your business is so bad you have to try to give your stuff away.
Does this count the number of Windows machines that were 'compromised' by BLASTER and its children?... More direct hacks on Linux machines might just mean that there was much more human effort expended.
If this is indeed the case, and I believe it is, the numbers show the opposite of what they would otherwise imply. I think everyone around here would agree attacks by real hackers reading the source code of every daemon you're running are much more difficult to defend against as oppose to the latest worm where your firewall just blocks port X that you don't really need anyhow and apply a patch for good measure. So if we're not counting automated attacks essentially what we're saying is it requires a hacker, not just another machine, to hack into a Linux server. To me, that says Linux in general is more difficult to hack.
From the article: The Sobig and MSBlast malware that afflict Microsoft platforms contributed significantly to the record estimate.
I find this a bit hard to swallow. Have they forgotten about Slammer? That one brought most of the net to a crawl. Is it possible that it compromised less than 4000 servers? And suppose a server is compromised twice, does that count as two or one in this count? I'm not even going to touch SoBig and MSBlaster.
What folks really want to know is how does OS choice affect security for their organization. This study doesn't give them that information.
1) You need to get a sense of reporting bias.
2) you need to make sure you are comparing
servers in similar situations
(i.e. Linux servers at major, unpopular
corporations vs. Windows servers at major,
unpopular corporations)--and make sure they
are equally interesting targets.
I can believe that ISP's that service
certain neighborhoods are especially vulnerable
to attack--and that ISP's don't use Windows.
3) I would compare how setting affects this. I
could believe for example that Linux/BSD
are much more secure in the hands of
a professional and Linux is less secure in the
hands of a novice.
I have heard of VERY few people running Apache on Windows. What's the point?
It would be stupid and reckless to tell a bunch of MCSE's to scrap a Windows server and replace it with Linux. If your organization doesn't have any Linux experience, the next best thing to moving away from Windows is using Apache instead of IIS.
Come on, where do they get these figures? In August alone:
From NetworkWoldFusion
The Blaster worm - also known as MSBlast or LoveSAN - has spread rapidly since it was first noticed on Monday. It has infected an estimated 188,000 systems running Microsoft operating systems, including Windows XP, Windows 2000, Windows 2003 and NT, that are unpatched for the so-called RPC vulnerability discovered last month, according to a security firm tracking the worm.
They didn't count them. Why? Most of them aren't servers, right? Well how did they differentiate Linux servers then? I bet they didn't -- did they check and only record RH Advanced Server and disregard all the RH Workstation. I doubt it. This is pure FUD by a place that has trouble with math.
To add to that, there are often G&M links on /.. It's probably the Canadian equivalent of the New York Times.
Another thing that's not clear here is what is classified as a successful breach? Does that mean defacing a web page? Does that mean getting full access to the box? I've had a web page on my server get defaced because I forgot to upgrade PHP, but I didn't really care that much. On the other hand getting my box rooted by somebody is a serious problem.
This sig has been temporarily disconnected or is no longer in service
I'll probably get modded down for this, but oh well.
I post often about how Linux is no less insecure than Windows or any other OS. And constantly, I get bashed, downmodded, told that there are more Linux servers but are less hacked, etc.
And yet here is a study that shows otherwise. Now look at all those people try to dismiss it. Try to dance around it, making excuses, and so on. If this study had shown that Windows was the most breached, people would take it at face value and we'd have the requisite hundreds of "I told you so" posts, heresay, anecdotes from idiots who don't patch their servers, and so on.
I'm sorry, but I just wanted to say, I told you so. All operating systems are as secure as their admins. Microsoft has millions of dollars and some of the top programmers in the world. They're damn secure. So is Linux. So are all the others, reasonably speaking. Linux is not the end-all of secure systems, and this just makes people who act that way look like idiots (especially when they're making ridiculous excuses to try to diffuse the study).
"Sufferin' succotash."
I'm curious, was Slashdot afraid to put "Linux Most Breached Server?" in the headline? The stats were about most breached. The point wasn't who was most attacked. I guess that one word needed to be changed to soften the blow...
"Sufferin' succotash."
Similarly, of course, a Windows box that is not running IIS or any other Web server app is not susceptible to hacks that exploit those apps.
An insecure app does not make the underlying OS insecure.
I mean, certainly part of it could be attributed to the fact that there really aren't that many windows servers. And it's pretty obvious that he's not counting viri as hacks. Could be they're counting web site compromises as "Hacks", which would explain the numbers, but which is completely misleading. It's hard as hell to totally secure a website, but a properly configured webserver should default you to "Nobody" even if it gets hacked, which would allow a theoretical intruder little leeway.
Even so, I don't know about their numbers. I mean, hypothetically speaking, if I were to go after a server, would I try a linux box, which is at least middlin secure right out of the box, or would I go after an NT box, usually hopelessly insecure, admin'd by some silly MCSE who's probably not going to notice, and not going to be able to catch me, using any one of a number of common script tools to exploit any one of a number of massive windows security flaws?
I do security for at least part of my living, and I've always found windows to be laughably insecure. I broke the security on this one box 20 TIMES in 2 weeks, and every time it was a new flaw. (And a new check for me. Mmmmmmm. Windows money.) I've never managed to do that to a linux box.
I'd really want to see their data. I mean, sure you can crack a linux box, but the easiest ways are 1) Social engineering and 2) physical access.
I've had MS guys give me their admin passwords OVER the fricking phone.
Bah. Anyway. This sounds like FUD to me.
ad logicam Claiming a proposition is false because it was presented as the conclusion of a fallacious argument.
> As to web sites they *appear* to count each
> web site affected. So a single linux breakin
> on a big hosting site scores 10,000 while
> nobody hosts 10,000 sites on a windows box.
So, how does this mitigate the damage? If someone cracks a Linux box hosting 10,000 web sites -- well, then 10,000 web sites have now been compromised. The impact is just as significant.
In the case of the "lazy" admin, I've watched how over worked MS-only shops become - patches often break things, fail to fix what they claim to, or (re-)introduce additional exploits and therefore must be tested very thoroughly before going onto a production system. Some shops try to save money and have only one server, thus they pay big time for mistakes...
In those cases, breaches are due to the patches themselves breaking things or not working. Can you say NT sp2?
Beta is broken and the link to classic doesn't work. Stop wasting our time or there won't be anybody left here.
The title states 'Attacked', but the article talks about successful compromises (Hacked).
Linux is the most hacked system?
Hmm...
Seems like a bit of reader manipulation there Slashdot!