Re: [gentoo-user] Curious pattern in log files from ssh...
On Fri, Dec 5, 2008 at 10:05 AM, Evgeniy Bushkov <[EMAIL PROTECTED]> wrote: > Adam Carter пишет: >>> >>> Also take a note that there are no "known-compromised hosts" >>> >> >> What about hosts listed in RBLs? >> http://en.wikipedia.org/wiki/Comparison_of_DNS_blacklists. It would be >> interesting to see if how much correlation there is between ssh brute >> forcing bots and the contents of the various lists. >> > > It's just interesting. But I don't trust them enough. I don't know how these > lists were composed. We've periodically seen viruses outbreaks, some > computers IPs could get into lists because of trojans and so on. One day you > won't reach your server from your own home computer... The fact that a lot of 'compromised hosts' are home users with providers like comcast, verizon, etc lends another trouble as well... dynamic IPs mean that the next person with the luck of the draw in getting that IP can't reach your servers either, and if *you* happen to be that person, no reasonable whitelist will ever get you back in from that location until you get another IP. >> >> >>> >>> because ANY IP can be forged. >>> >> >> Its easy enough to forge a SYN, but to setup a session so you can make a >> password guessing attempt requires that you also get the packets back from >> the server, which is an order of magnitude more difficult. Ever since OSes >> have implemented well chosen initial sequence numbers, spoofing of TCP >> sessions has become very difficult. >> >> > > I agree but as admin I prefer to think about many things worse than they > really are. If something wrong is possible it's better to avoid it > beforehand. > > Best regards, > Evgeniy B. Careful with that line of thinking... you'll inevitably come to the conclusion that there's no hope and you're better off just turning the system off, unplugging it from the wall, and locking it into a very sturdy vault deep beneath a very solid mountain! (until you ponder yourself insane over the security risks that exist even then, let alone the impact on usability) -- Poison [BLX] Joshua M. Murphy
Re: [gentoo-user] Curious pattern in log files from ssh...
Adam Carter пишет: Also take a note that there are no "known-compromised hosts" What about hosts listed in RBLs? http://en.wikipedia.org/wiki/Comparison_of_DNS_blacklists. It would be interesting to see if how much correlation there is between ssh brute forcing bots and the contents of the various lists. It's just interesting. But I don't trust them enough. I don't know how these lists were composed. We've periodically seen viruses outbreaks, some computers IPs could get into lists because of trojans and so on. One day you won't reach your server from your own home computer... because ANY IP can be forged. Its easy enough to forge a SYN, but to setup a session so you can make a password guessing attempt requires that you also get the packets back from the server, which is an order of magnitude more difficult. Ever since OSes have implemented well chosen initial sequence numbers, spoofing of TCP sessions has become very difficult. I agree but as admin I prefer to think about many things worse than they really are. If something wrong is possible it's better to avoid it beforehand. Best regards, Evgeniy B. smime.p7s Description: S/MIME Cryptographic Signature
Re: [gentoo-user] Curious pattern in log files from ssh...
Alan McKinnon wrote: > On Thursday 04 December 2008 21:03:17 Christian Franke wrote: > >> I just don't see what blocking ssh-bruteforce attempts should be good >> for, at least on a server where few _users_ are active. >> > Two reasons: > > a. Maybe, just maybe, you overlooked something. Belts, braces and a > drawstring > for good measure is not a bad thing. > > b. You probably want to get all that crap out of your log files off into some > other place where you can cope with it. Parsing auth log files that are 95% > brute force attempts is no fun. I like to have the crap in place A and the > real stuff in place B, makes my job so much easier > I agree 100% with the above - another issue is that I'd like to block all traffic from malicious hosts - I realise that the traffic is low at the moment, but that need not be the case in future. >> Also, things like fail2ban add new attack-possibilities to a system, I >> remember the old DoS for fail2ban, resulting from a wrong regex in log >> file parsing, but I think at least this is fixed now. >> > Whereas that is true enough in itself, the actual risk of such is rather low > in comparison to the gains. Hence it is not a valid reason to not use > fail2ban and such-like apps. The issue for me is that the cost of a DOS is far, far lower than the cost of a break-in. The cost of a DOS that prevents access from new hosts is orders of magnitude lower than the cost of a DOS. Everyone's risk profiles are different - but, for me, keeping out intruders is critical (they may result in unrecoverable data loss) and my accessibility objective is that it be the 'norm' that I can log in with an unusual-username and complex password from a trustworthy PC whose IP address can not be determined in advance... using only bog-standard tools and no non-remembered personal data. I'm coming around to the idea of port-knocking, but my gut instinct is that it is a bit baroque and has potential for me to louse-up its implementation... It definitely adversely affects usability - though, I admit, less than I first suspected. I'm still quite interested in the idea of identifying botnets where used to subvert the tactics used by fail2ban; blacklist.py, etc. and using these to, in turn, block access to any service... including, for example, hosted web-services which are, potentially, in spite of taking all the obvious precautions, more vulnerable to attack - IMHO. I'm definitely thinking that it would be a good idea if there were a way to publish botnet lists... such that they could be collated and turned into a DNSBL style resource. If such a resource existed, I'd definitely chose to use it (overridden by a few whitelist entries of my own - just-in-case...) and I'd be very happy to report back to it in order to help keeping this problem under control. Incidentally, I'd also consider it useful to monitor this block list for any occurrence of my own IP address - since that would be an early indication that one of my hosts may be compromised.
Re: [gentoo-user] Curious pattern in log files from ssh...
On Thursday 04 December 2008, Steve wrote: > Simon wrote: > > Since it is very unlikely that the attacker is targeting you > > specifically, changing the port number (and removing root access) will > > very likely stop the attack forever. Though, if the attacker did > > target you, then you will need some more security tools (intrusion > > detection, etc...). > > I recognise that this doesn't seem to be a targeted attack - but it is > still frustrating to find that someone has evaded my IP blocking > strategy... even though they pose only a slightly elevated risk by > having done so. (Of course, I don't permit root login - that would be > madness... and, as far as I'm aware, no-one has guessed even a valid > user name... they're all obscure!) > > The thing that strikes me is that, in evading my blocking strategy, they > clearly identified a bot-net of compromised hosts. With this in mind, > ideally, I'd like to: > > 1. Automatically detect and block all future attacks on all ports from > all hosts which are involved in this coordinated attack. These hosts > can't be trusted not to be malicious. > 2. Somehow inform the administrator of the hosts attacking me (in a > respectful way) since, I presume, they are unaware that their host is > involved in the attack. > 3. Ideally, share this kind of information so that myself and others are > better protected from bot-net attacks in future. > > It's the sort of thing I imagine has already been done - and there's no > point in re-inventing the wheel. I recall something similar whereby the attacked machines would automatically launch an attack on the botnet/spammer to effect a DoS. Then the spammers complained and the guys who had written the software were forced by the police to recall it . . . sometimes I wonder. Anyway, I'm a bit thin on details - this was all the rage about 4-5 years ago as a legit way to defend yourself against spam. What I think is required is a script which will identify the compromised machine and promptly reformat its MSWindows OS - problem solved. Of course how you keep tabs on this tool not being misused is another thing. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Curious pattern in log files from ssh...
On December 4, 2008, Adam Carter wrote: > > Open a Wiki page on Wikipedia, update it every so often and > > provide simple > > parser for it so others can recycle same IPs. Since it's a > > Wiki page - others > > can update it as well (including botnet owners, but then > > they'd have to reveal themselves - tricky situation) :) > > Reveal themselves in what way? If you're taking about source IP, they can > just use one of their bots to make the page update... true. -- Dmitry Makovey Web Systems Administrator Athabasca University (780) 675-6245 signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Curious pattern in log files from ssh...
Dmitry S. Makovey wrote: On December 3, 2008, Steve wrote: Dmitry S. Makovey wrote: well. Nobody but you knows your requiremens and specifics - we're just listing options. It's up to you to either take 'em or leave 'em ;) Fair enough - but I've still not found an option for sharing/using shared block lists for bot-nets. Open a Wiki page on Wikipedia, update it every so often and provide simple parser for it so others can recycle same IPs. Since it's a Wiki page - others can update it as well (including botnet owners, but then they'd have to reveal themselves - tricky situation) :) I hear the botnet owners have 1 or 2 spare machine scattered around the world they can proxy through... :) Shawn
RE: [gentoo-user] Curious pattern in log files from ssh...
> > Also take a note that there are no "known-compromised hosts" > > What about hosts listed in RBLs? > http://en.wikipedia.org/wiki/Comparison_of_DNS_blacklists. It > would be interesting to see if how much correlation there is > between ssh brute forcing bots and the contents of the various lists. Maybe http://wiki.duskglow.com/tiki-index.php?page=Packetbl "PacketBL is a program that uses DNS blocklists to determine whether to accept or reject packets" Used with dnsbl.ahbl.org "Aggregate zone, contains UCE/bulk email senders, open proxies, open relays, trojaned/infected machines, comment/trackback spammers" would be a good solution.
RE: [gentoo-user] Curious pattern in log files from ssh...
> Open a Wiki page on Wikipedia, update it every so often and > provide simple > parser for it so others can recycle same IPs. Since it's a > Wiki page - others > can update it as well (including botnet owners, but then > they'd have to reveal themselves - tricky situation) :) Reveal themselves in what way? If you're taking about source IP, they can just use one of their bots to make the page update...
RE: [gentoo-user] Curious pattern in log files from ssh...
> Also take a note that there are no "known-compromised hosts" What about hosts listed in RBLs? http://en.wikipedia.org/wiki/Comparison_of_DNS_blacklists. It would be interesting to see if how much correlation there is between ssh brute forcing bots and the contents of the various lists. > because ANY IP can be forged. Its easy enough to forge a SYN, but to setup a session so you can make a password guessing attempt requires that you also get the packets back from the server, which is an order of magnitude more difficult. Ever since OSes have implemented well chosen initial sequence numbers, spoofing of TCP sessions has become very difficult.
Re: [gentoo-user] Curious pattern in log files from ssh...
On Thursday 04 December 2008 21:03:17 Christian Franke wrote: > On 12/03/2008 09:02 PM, Steve wrote: > > I've recently discovered a curious pattern emerging in my system log > > with failed login attempts via ssh. > > > > I'm not particularly concerned - since I'm confident that all my users > > have strong passwords... but it strikes me that this data identifies a > > bot-net that is clearly malicious attempting to break passwords. > > > > Sure, I could use IPtables to block all these bad ports... or... I could > > disable password authentication entirely... but I keep thinking that > > there has to be something better I can do... any suggestions? Is there > > a simple way to integrate a block-list of known-compromised hosts into > > IPtables - rather like my postfix is configured to drop connections from > > known spam sources from the sbl-xbl.spamhaus.org DNS block list, for > > example. > > I just don't see what blocking ssh-bruteforce attempts should be good > for, at least on a server where few _users_ are active. Two reasons: a. Maybe, just maybe, you overlooked something. Belts, braces and a drawstring for good measure is not a bad thing. b. You probably want to get all that crap out of your log files off into some other place where you can cope with it. Parsing auth log files that are 95% brute force attempts is no fun. I like to have the crap in place A and the real stuff in place B, makes my job so much easier > > The chance that security of a well configured system will be compromised > by that is next to zero, and on recent systems it is also impossible to > cause significant load with ssh-login-attempts. Uh-huh. We all said that for many years. Then some bright spark actually looked at the patches the debian openssh maintainer was applying and we all had one of those special "oops..." moments Did you have any idea of just how weak certs made on a debian box were before it hit the headlines? No-one I know did. > Also, things like fail2ban add new attack-possibilities to a system, I > remember the old DoS for fail2ban, resulting from a wrong regex in log > file parsing, but I think at least this is fixed now. Whereas that is true enough in itself, the actual risk of such is rather low in comparison to the gains. Hence it is not a valid reason to not use fail2ban and such-like apps. If it were, we should all just stop using iptables and libwrap and openssl on the off-chance that maybe, just maybe, they open an attack vector. But that's silly reasoning right? -- alan dot mckinnon at gmail dot com
Re: [gentoo-user] Curious pattern in log files from ssh...
On December 4, 2008, Christian Franke wrote: > I just don't see what blocking ssh-bruteforce attempts should be good > for, at least on a server where few _users_ are active. Considering how much creative paranoia I've exposed in this thread it might come as a surprise, but I do agree with the above statement. Strong passwords (or key-only authentication) would prevent brute-force attacks from being successfull. The only thing that is semi-usefull side-effect is that you can identify compromised machines and deny ANY type of traffic from them preventing possible DoS launched against you. But then IPs are so easy to spoof :) Balance is what makes sysadmin comfortable enough and doesn't compromise usability of the server, so everybody decides for themselves. OP obviously wants that "extra" layer of protection and notification so with a bit of creativity and some external tools it's possible to achieve. As long as he doesn't forget about other aspects of security - he should do just fine with all those extra measures :) -- Dmitry Makovey Web Systems Administrator Athabasca University (780) 675-6245 signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Curious pattern in log files from ssh...
On 12/03/2008 09:02 PM, Steve wrote: > I've recently discovered a curious pattern emerging in my system log > with failed login attempts via ssh. > > I'm not particularly concerned - since I'm confident that all my users > have strong passwords... but it strikes me that this data identifies a > bot-net that is clearly malicious attempting to break passwords. > > Sure, I could use IPtables to block all these bad ports... or... I could > disable password authentication entirely... but I keep thinking that > there has to be something better I can do... any suggestions? Is there > a simple way to integrate a block-list of known-compromised hosts into > IPtables - rather like my postfix is configured to drop connections from > known spam sources from the sbl-xbl.spamhaus.org DNS block list, for > example. I just don't see what blocking ssh-bruteforce attempts should be good for, at least on a server where few _users_ are active. The chance that security of a well configured system will be compromised by that is next to zero, and on recent systems it is also impossible to cause significant load with ssh-login-attempts. Also, things like fail2ban add new attack-possibilities to a system, I remember the old DoS for fail2ban, resulting from a wrong regex in log file parsing, but I think at least this is fixed now. Regards, Christian Franke signature.asc Description: OpenPGP digital signature
Re: [gentoo-user] Curious pattern in log files from ssh...
On December 3, 2008, Steve wrote: > Dmitry S. Makovey wrote: > >> Erm - surely I either need to set up my client to port-knock... which > >> is a faff I'd rather avoid... in order to use the technique. > > > > nope. just start connection. wait a minute. cancel. start another one. > > wait a minute. cancel. start new one - voila! :) > > Eeew... especially as this would apply to all connections - even the > ones where I have a DSA key. I might be able to cope with this if it > only applied to my initial connection, from which I could grab a copy of > the DSA key. Ok, let's theoreticise some more. My paranoia feels particularly frisky today, so here it is: remember, I've mentioned origianlly that once you authenticate successfully once with DSA key - your IP is whitelisted. So subsequent connections go right through. > > well. Nobody but you knows your requiremens and specifics - we're just > > listing options. It's up to you to either take 'em or leave 'em ;) > > Fair enough - but I've still not found an option for sharing/using > shared block lists for bot-nets. Open a Wiki page on Wikipedia, update it every so often and provide simple parser for it so others can recycle same IPs. Since it's a Wiki page - others can update it as well (including botnet owners, but then they'd have to reveal themselves - tricky situation) :) P.S. I think I'd better stop with my mad science projects here before I go too far and invent brand new theory on host protection ;) -- Dmitry Makovey Web Systems Administrator Athabasca University (780) 675-6245 signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Curious pattern in log files from ssh...
Simon wrote: > Since it is very unlikely that the attacker is targeting you > specifically, changing the port number (and removing root access) will > very likely stop the attack forever. Though, if the attacker did > target you, then you will need some more security tools (intrusion > detection, etc...). I recognise that this doesn't seem to be a targeted attack - but it is still frustrating to find that someone has evaded my IP blocking strategy... even though they pose only a slightly elevated risk by having done so. (Of course, I don't permit root login - that would be madness... and, as far as I'm aware, no-one has guessed even a valid user name... they're all obscure!) The thing that strikes me is that, in evading my blocking strategy, they clearly identified a bot-net of compromised hosts. With this in mind, ideally, I'd like to: 1. Automatically detect and block all future attacks on all ports from all hosts which are involved in this coordinated attack. These hosts can't be trusted not to be malicious. 2. Somehow inform the administrator of the hosts attacking me (in a respectful way) since, I presume, they are unaware that their host is involved in the attack. 3. Ideally, share this kind of information so that myself and others are better protected from bot-net attacks in future. It's the sort of thing I imagine has already been done - and there's no point in re-inventing the wheel.
Re: [gentoo-user] Curious pattern in log files from ssh...
Steve пишет: I've recently discovered a curious pattern emerging in my system log with failed login attempts via ssh. Previously, I noticed dictionary attacks launched - which were easy to detect... and I've a process to block the IP address of any host that repeatedly fails to authenticate. What I see now is quite different... I'm seeing a dictionary attack originating from a wide range of IP addresses - testing user-names in sequence... it has been in progress since 22nd November 2008 and has tried 7195 user names in alphabetical order from 521 distinct hosts - with no successive two attempts from the same host. I'm not particularly concerned - since I'm confident that all my users have strong passwords... but it strikes me that this data identifies a bot-net that is clearly malicious attempting to break passwords. Sure, I could use IPtables to block all these bad ports... or... I could disable password authentication entirely... but I keep thinking that there has to be something better I can do... any suggestions? Is there a simple way to integrate a block-list of known-compromised hosts into IPtables - rather like my postfix is configured to drop connections from known spam sources from the sbl-xbl.spamhaus.org DNS block list, for example. Break in attempts today (attempted username/IP address): -- Hi. Best of all you can add iptables rules. It's better then use any script. Also take a note that there are no "known-compromised hosts" because ANY IP can be forged. I've sometimes seen such rules in the internet. These I use in my firewall: iptables -A INPUT -i eth0 -p tcp -m state --state NEW --dport 22 -m recent --name sshattack --set iptables -A INPUT -i eth0 -p tcp -m state --state NEW --dport 22 -m recent --name sshattack --rcheck --seconds 60 --hitcount 4 -j LOG -m limit --limit 3/minute --limit-burst 3 --log-level 4 --log-prefix 'SSH REJECT: ' iptables -A INPUT -i eth0 -p tcp -m state --state NEW --dport 22 -m recent --name sshattack --rcheck --seconds 60 --hitcount 4 -j REJECT --reject-with tcp-reset These rules give you possibility to use your ssh service from any IP but reject repeated login attempts from malicious users. You can tune parameter --hitcount to limit amount of sequential login attempts per minute. Also you can look at /proc/net/ipt_recent/sshattack for malicious IPs and how often they were used. Best regards, Evgeniy B. smime.p7s Description: S/MIME Cryptographic Signature
Re: [gentoo-user] Curious pattern in log files from ssh...
I noticed the same thing on my host several weeks ago. I strongly suggest removing root access to your ssh, root is probably being tried by more than 50% of all login attempts... the other trials are semi-intelligent random usernames (ie, users that might really well exists, like 'apache' etc... but other usernames which may not like 'albert'). If your username is not part of the list of attempts, then it won't be tried much, and I once found out that if your password is alphanumeric with lower and upper cases, the hacker as a worst chance of finding your password in (26*2+10)^8(chars long) = 62^8 = 2.18e14 steps or 218 millions of millions of steps. This is assuming they try the correct username each time! The other thing you should do is place ssh on another port, very high. IIRC, port numbers are 16bits and can go as high as 65k... you could use 22xxx where xxx is a random favorite number for example. Since it is very unlikely that the attacker is targeting you specifically, changing the port number (and removing root access) will very likely stop the attack forever. Though, if the attacker did target you, then you will need some more security tools (intrusion detection, etc...). Good luck! Simon Steve wrote: I've recently discovered a curious pattern emerging in my system log with failed login attempts via ssh. Previously, I noticed dictionary attacks launched - which were easy to detect... and I've a process to block the IP address of any host that repeatedly fails to authenticate. What I see now is quite different... I'm seeing a dictionary attack originating from a wide range of IP addresses - testing user-names in sequence... it has been in progress since 22nd November 2008 and has tried 7195 user names in alphabetical order from 521 distinct hosts - with no successive two attempts from the same host. I'm not particularly concerned - since I'm confident that all my users have strong passwords... but it strikes me that this data identifies a bot-net that is clearly malicious attempting to break passwords. Sure, I could use IPtables to block all these bad ports... or... I could disable password authentication entirely... but I keep thinking that there has to be something better I can do... any suggestions? Is there a simple way to integrate a block-list of known-compromised hosts into IPtables - rather like my postfix is configured to drop connections from known spam sources from the sbl-xbl.spamhaus.org DNS block list, for example. Break in attempts today (attempted username/IP address): -- huck 190.60.41.82 huckleberry 81.196.122.2 huckleberry 58.39.145.213 huckleberry 60.230.184.143 hue 58.196.4.2 hue 83.228.92.228 huela 193.41.235.225 huela 193.41.235.225 huey 201.21.216.198 huey 81.149.101.27 hugh 200.123.174.145 hugh 83.228.92.228 hugh 212.46.24.146 hugo 195.234.169.138 hugo 193.86.111.6 hugo 201.224.199.201 hume 69.217.30.214 hume 80.118.132.88 hummer 71.166.159.177 hummer 200.126.119.91 hummer 61.4.210.33 humphrey 80.34.55.88 humphrey 213.163.19.158 humvee 85.222.53.48 humvee 80.24.4.23 hung 61.47.31.130 hung 70.46.140.187 hunter 67.40.86.204 hunter 83.228.92.228 hunter 200.60.156.90 huong 207.250.220.196 huong 125.63.77.3 huong 200.62.142.212 huslu 219.93.187.38 huslu 121.223.228.249 huslu 200.29.135.50 hussein 200.60.156.90 hussein 200.6.220.46 hussein 125.63.77.3 huy 60.191.111.234 huy 200.79.25.39 huyen 213.136.105.130 huyen 190.144.61.58 huyen 121.33.199.37 hy 121.33.199.37 hy 90.190.96.46 hyacinth 81.196.122.2 hyacinth 189.43.21.244 hyacinth 99.242.205.242 hyman 201.21.216.198 hypatia 218.28.143.246 hypatia 195.234.169.138 iain 200.118.119.48 iain 124.42.124.87 iain 194.224.118.61 ian 189.56.92.42 ian 201.28.119.60 ian 210.187.18.199 ianna 211.154.254.120 ianna 84.242.66.10 ianna 193.41.235.225 ianthe 81.246.26.179 ibtesam 87.30.163.87 ichabod 201.251.61.108 ida 62.61.141.93 ida 80.24.4.23 idalee 85.222.53.48 idalee 190.144.61.58 --
Re: [gentoo-user] Curious pattern in log files from ssh...
Dmitry S. Makovey wrote: >> Erm - surely I either need to set up my client to port-knock... which >> is a faff I'd rather avoid... in order to use the technique. > nope. just start connection. wait a minute. cancel. start another one. wait a > minute. cancel. start new one - voila! :) > Eeew... especially as this would apply to all connections - even the ones where I have a DSA key. I might be able to cope with this if it only applied to my initial connection, from which I could grab a copy of the DSA key. > well. Nobody but you knows your requiremens and specifics - we're just > listing > options. It's up to you to either take 'em or leave 'em ;) Fair enough - but I've still not found an option for sharing/using shared block lists for bot-nets.
Re: [gentoo-user] Curious pattern in log files from ssh...
On December 3, 2008, Steve wrote: > Paul Hartman wrote: > > I think using Dmitry's idea of rejecting the first 2 connections, but > > then allowing it as normal on the third attempt would satisfy your > > requirements for being on the normal port, allowing all IPs and > > requiring no special setup on the client end (other than knowing they > > have to to retry twice). > > Erm - surely I either need to set up my client to port-knock... which is > a faff I'd rather avoid... in order to use the technique. nope. just start connection. wait a minute. cancel. start another one. wait a minute. cancel. start new one - voila! :) > While I recognise port knocking as a valuable strategy in some > circumstances, it seems a very bad fit for my needs. well. Nobody but you knows your requiremens and specifics - we're just listing options. It's up to you to either take 'em or leave 'em ;) -- Dmitry Makovey Web Systems Administrator Athabasca University (780) 675-6245 signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Curious pattern in log files from ssh...
Paul Hartman wrote: > I think using Dmitry's idea of rejecting the first 2 connections, but > then allowing it as normal on the third attempt would satisfy your > requirements for being on the normal port, allowing all IPs and > requiring no special setup on the client end (other than knowing they > have to to retry twice). > Erm - surely I either need to set up my client to port-knock... which is a faff I'd rather avoid... in order to use the technique. Port knocking would be especially infuriating from trusted clients where I'd like to use standard software like WinSCP; Putty; Symbian Putty - etc. While I recognise port knocking as a valuable strategy in some circumstances, it seems a very bad fit for my needs. GEO-IP blocking would be fairly good... if I could limit this to password authentication only - as would blacklisting known bot-net participants. While these exotic ideas are interesting - a better way to identify malicious hosts is, by far, my preferred solution.
Re: [gentoo-user] Curious pattern in log files from ssh...
On December 3, 2008, Paul Hartman wrote: > Of course, this is assuming the botnet stops after rejected connections... oh no, not rejected - dropped ;) let them go through pains of timing out without knowing if anything is actually listening on the other side ;) -- Dmitry Makovey Web Systems Administrator Athabasca University (780) 675-6245 signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Curious pattern in log files from ssh...
On Wed, Dec 3, 2008 at 4:55 PM, Steve <[EMAIL PROTECTED]> wrote: > Dmitry S. Makovey wrote: >> P.S. I actually don't do any of the above. It was just a surge of creative >> paranoia >> in response to initial request :) > All good ideas - except selling the blacklist... I'd be happiest to > share my blacklist for free... my objective is to minimise exposure to > botnets - rather than to accept another level of complexity with > legitimate use. I think using Dmitry's idea of rejecting the first 2 connections, but then allowing it as normal on the third attempt would satisfy your requirements for being on the normal port, allowing all IPs and requiring no special setup on the client end (other than knowing they have to to retry twice). Of course, this is assuming the botnet stops after rejected connections...
Re: [gentoo-user] Curious pattern in log files from ssh...
Dmitry S. Makovey wrote: > P.S. I actually don't do any of the above. It was just a surge of creative > paranoia > in response to initial request :) All good ideas - except selling the blacklist... I'd be happiest to share my blacklist for free... my objective is to minimise exposure to botnets - rather than to accept another level of complexity with legitimate use.
RE: [gentoo-user] Curious pattern in log files from ssh...
> I previously used denyhosts - but (I can't remember why) it became > preferable to block with IPtables rather than with > tcpwrappers... which > prompted me to dump it in favour of a bespoke script based upon > blacklist.py (http://blinkeye.ch/mediawiki/index.php/SSH_Blocking) - > though, now, I'm tempted by the more professional looking sshguard - > thanks for the tip. Of course, this doesn't really address > the problem > I posted about - because I'm now faced with a highly distributed > dictionary attack... Fail2ban is iptables based. From the website it now appears to have a map feature so if say you notice most of the attacks coming from China, and none of you ssh useres are in China, you could perhaps block the entire country with http://people.netfilter.org/~peejix/geoip/howto/geoip-HOWTO.html
Re: [gentoo-user] Curious pattern in log files from ssh...
On December 3, 2008, Steve wrote: > I have, in the past, used DSA only keys - but this was frustrating on > several occasions when I wanted access to my server and didn't have my > SSH keys available to me... I almost always connect using a key pair > rather than a password - but the password option is very useful to allow > me to get hold of my SSH keys in the first place in some environments. > If I found a distributed attack on a valid user name, for example, I'd > consider this a critical change - however inconvenient. get yourself some portable linux device capable of either USB, ethernet or wifi connection (OpenMoko, Nokia NXXX, etc.) plug your keys there - and voila, you've got yourelf both secure terminal and key storage in one box. I would be highly suspicious initiating SSH connection with my servers from untrusted box (which is any box not built and maintained by me ;) ) as there is a chance of keylogger (no matter how friendly owner of spoken box is - you don't know if he wasn't hacked and you have no time for even casual checking). You can use variation of port-knocking and reverse your strategy based on the pattern: 1. drop first connection from specified IP and record it in "first_try" table 2. drop second connection from specified IP and record it in "second_try" table 3. if IP is in both first_try and second_try - allow it to attempt authentication but only with the keys. (removing it from *_try tables and possibly recording it in whitelist) 4. if IP fails X number of attempts within specified timeframe - remove from whitelist and record in blacklist bit tricky logic, but fairly simple to implement (I use *BSD PF so no ready recipe for iptables here ;) ). bit paranoid, but it covers your initial concern with distributed attack and single-attempts. You can further collect older entries from first_try into blacklist and do whatever you please with them. You can also collect high-frequency attempts into blacklist and have very big blacklist you can sell off on eBay :) P.S. I actually don't do any of the above. It was just a surge of creative paranoia in response to initial request :) -- Dmitry Makovey Web Systems Administrator Athabasca University (780) 675-6245 signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Curious pattern in log files from ssh...
Thanks for all the replies so far... I'll reply once to these... (Oh, and when I said "ports" in my original post, I meant "addresses" - my typing fingers just ignored my brain...) I'm against a 'novel port' approach - as I am against port-knocking (for my server) because these may prove challenging for the environments from which I may want to log on. I want to retain a 'standard' service to make it easiest for me to connect to my server from a remote site without requiring reconfiguration of firewalls etc. I have, in the past, used DSA only keys - but this was frustrating on several occasions when I wanted access to my server and didn't have my SSH keys available to me... I almost always connect using a key pair rather than a password - but the password option is very useful to allow me to get hold of my SSH keys in the first place in some environments. If I found a distributed attack on a valid user name, for example, I'd consider this a critical change - however inconvenient. I previously used denyhosts - but (I can't remember why) it became preferable to block with IPtables rather than with tcpwrappers... which prompted me to dump it in favour of a bespoke script based upon blacklist.py (http://blinkeye.ch/mediawiki/index.php/SSH_Blocking) - though, now, I'm tempted by the more professional looking sshguard - thanks for the tip. Of course, this doesn't really address the problem I posted about - because I'm now faced with a highly distributed dictionary attack... It strikes me that, given the conclusive nature of this attack (which, by virtue of the fact that the usernames are attempted in alphabetical order proves it to be a single coordinated attack) I can create a list of a large number of IP addresses - which, likely, correspond to compromised hosts. It strikes me that this would be a perfect source of information to set up a block list... and, if others' logs show similar attacks, it should be easy enough to combine this data to provide distributed protection from a distributed attack. I don't think for one second that this attack is targeted - neither my hardware or the information on my server is particularly interesting to anyone but me. It would be extremely interesting to me, however, if it were to transpire that my IP address originated login attempts such as these - as this would clearly demonstrate it to be compromised... I suspect, too, the ISPs should be interested to inform their subscribers in the interest of security... though, of course, I recognise that this is being optimistic. When I exposed my server to internet SSH logins, I carefully considered security... though I also had to consider convenience - since that was the only reason for doing so in the first place. If I could block all IPs suspected of being in a bot-net - then this would be an improvement in security without a great cost in terms of lost convenience. Right now, in the context of this attack which circumvents my earlier blocking strategy, I'm looking for a viable blacklist solution in order to avoid white-listing. A potential solution for me would be to have sshd be far more choosy about source IPs when using password authentication... for example, restricting it to hosts in the UK... but still allowing remote access wherever I've propagated DSA keys... but I think this would be tricky to set up. A shared block-list, I suspect, would be the most effective response to this attack... and the response most likely to minimise others' exposure, too. Steve
Re: [gentoo-user] Curious pattern in log files from ssh...
On December 3, 2008, Steve wrote: > Sure, I could use IPtables to block all these bad ports... or... I could > disable password authentication entirely... but I keep thinking that > there has to be something better I can do... any suggestions? Is there > a simple way to integrate a block-list of known-compromised hosts into > IPtables - rather like my postfix is configured to drop connections from > known spam sources from the sbl-xbl.spamhaus.org DNS block list, for > example. I went the path of paswordless entries (i.e. DSA/RSA keys) and I think it helped a lot, no botnet/worm/cracker is known to do selective key assembly so far and it's a labour-intensive process. I think applying keys is a very good step forward (well, and make sure every externally exposed service is properly patched and secured ;) ). -- Dmitry Makovey Web Systems Administrator Athabasca University (780) 675-6245 signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Curious pattern in log files from ssh...
On Wednesday 03 December 2008 22:02:43 Steve wrote: > I've recently discovered a curious pattern emerging in my system log > with failed login attempts via ssh. > > Previously, I noticed dictionary attacks launched - which were easy to > detect... and I've a process to block the IP address of any host that > repeatedly fails to authenticate. > > What I see now is quite different... I'm seeing a dictionary attack > originating from a wide range of IP addresses - testing user-names in > sequence... it has been in progress since 22nd November 2008 and has > tried 7195 user names in alphabetical order from 521 distinct hosts - > with no successive two attempts from the same host. Slashdot yesterday, read the front page It seems to be a co-ordinated and very well synchronized stealth bot-net. You are one of many that has noticed this. I am noticing scans on machines that have never been scanned before in all the time they have been up. You should indeed be very concerned and take extra special due care with your security arrangements currently. In fact, if you admin machines that are in any way critical, you really *really* should be undertaking a thorough security audit and make very sure you have done everything and covered all your bases. -- alan dot mckinnon at gmail dot com
Re: [gentoo-user] Curious pattern in log files from ssh...
On Wed, Dec 3, 2008 at 2:02 PM, Steve <[EMAIL PROTECTED]> wrote: > I've recently discovered a curious pattern emerging in my system log > with failed login attempts via ssh. > > Previously, I noticed dictionary attacks launched - which were easy to > detect... and I've a process to block the IP address of any host that > repeatedly fails to authenticate. > > What I see now is quite different... I'm seeing a dictionary attack > originating from a wide range of IP addresses - testing user-names in > sequence... it has been in progress since 22nd November 2008 and has > tried 7195 user names in alphabetical order from 521 distinct hosts - > with no successive two attempts from the same host. This has been going on all year, you're lucky if you just started getting it. :) AFAIK nobody has found any specific fingerprint or anything to block it by. The "solution" seems to be: only allow SSH from specific IP addresses, don't use port 22, don't use password auth, use some kind of portknocking, etc. as you already alluded to. If you Google for distributed ssh brute force attacks, there are some fairly detailed articles out there from earlier in the year. Good luck :) Paul
Re: [gentoo-user] Curious pattern in log files from ssh...
You could try sshguard or denyhosts.
