Access Granted

Back in April, I discussed the March breach in security at RSA. After I posted my article, I was contacted by a friend of mine that works for a certain “Three Letter Acronym” agency whose name and organization will remain un-named. My friend asked me if I might happen to know exactly what was compromised at RSA? While I couldn’t be certain, I suggested what I (and maybe most of you) had already heard via news channels, media, and security contacts. While I know more details on what actually happened, I don’t know specifics. From my knowledge, what was known to be compromised was possibly the algorithm seed, and maybe the key generation time. The algorithm itself, has been publically known for years. From a security standpoint, this is like having a key, but not sure which lock it fits into.

I told my friend that there shouldn’t be too much to worry about, because while there’s enough pieces compromised to launch a brute force attack (basically, an educated guess that is retried until you succeed), there shouldn’t be enough to do a more damaging attack. Brute force attempts are fairly easy to spot, mitigate, and deny access. My friend was put, at least a little, at ease. That was until last week…

I received notification that sometime last week, a very large U.S defense contractor that uses SecureID tokens from RSA to provide two-factor authentication (something you have, and something you know; think of your Bank Card and your PIN) for remote VPN access to their corporate networks. Before Monday morning an alert went out, and all remote access to the internal corporate network was shutdown. Employees were notified that remote access could be down upwards of a week, possibly more. For telecommuters, this meant you either came into a branch office, or you simply could no longer work. Two days ago, my friend told me, a notification that every person who had an RSA SecureID token, would be getting a new one. This process, as I discussed in my earlier article, would take at least a few weeks to funnel out to everyone.

Along with this, all users (over 100,000 of them) would be required to change their passwords. The amount of help desk related issues this causes, simply means that administrative level files and access have almost assuredly been compromised.

From what I can tell, whomever hacked RSA, had now come into possession of the algorithm for the current tokens, and had then managed to install a key-stroke logger somewhere on the network. With both of these pieces, that unknown lock I discussed earlier in this article was now known.

While this was an expected outcome (most security folks like myself, have been awaiting such a breach), it was not enough to circumvent this from occurring. Shortly after the RSA breach became public knowledge, most companies that relied on SecureID for authentication, started requiring a second form of password before access to the network was granted. This, though as you can probably tell, would not resolve the issue if a key-logger was in place, as the hacker would know the password the minute it was typed.

I am a “Glass Half Full” sort of person, so I guess the silver lining in this story is that my friend and his staff were able to spot the intrusion, and acted appropriately to mitigate any further incidents. Kudos are warranted for such a feat as this is not an easy task. Although the aftershocks caused by this incident will be many, and far into the future.

While I am sure this is not an isolated incident, it is a major one, and one of the first public ones. At the time of this writing, I can state I know of others as well. This is not the first successful hacking attempt using the compromised SecureID technology. It certainly won’t be the last…

What concerns me most though, is that RSA has not been as forthwith in providing full disclosure about what was compromised and how vulnerable we are. RSA, if you’re listening, pretending like this didn’t happen and keeping it all secret, does not help things, it only makes them harder to track.

Even given all of the issues raised in this article, I don’t see anyone abandoning RSA or the SecureID product it sells. While most networks exchange token information over a secured and encrypted network path, this is only a false sense of security. My friend, and his organization can now attest to this.

If this can happen to a very secure network, employing some really talented security staff and products, it can happen to others as well. How far this will lead, and what sort of national secrets will be exposed now that such an attack has publically been proven to work, only time will tell.

Thus Quoth the Token Nevermore

Traditional tokens are a dying breed

There is an oft used term that a recent event is a 'sign of the times'. In 2011 though, technology has dictated the ringing of the bell announcing the dead - at least in terms of the traditional token. Gone are the days of the telltale SecurID key fob dangling from your house keys. The recent hacking of RSA exposed data that could ultimately compromise the security of the widely used SecurID token. In their defense, it's not the first time tokens have been attacked; and yes, compromised. Tokens are, in fact, still around. So, why, you may ask, is this year different from previous ones?

Let us review:

• During the 70's the 8 inch floppy was king




• The 80's gave way to the 5 inch floppy disk & later, the 3.5 inch




• The 90's took it a step further and gave us the CD, USB, and DVD for media alternatives




• All hail, the Blu-ray of the new millennium

What this shows is that all things change. Nothing, especially where technology is concerned, is constant. Two factor authentication is no exception. While two factor authentication has seen improvements since the days of yore with complex and cumbersome challenge tokens, we're still relying on this antiquated technology in the majority of physical tokens today.

If you think that a solution is great only because it's stood the test of time longer than most, think again. Columbus thought that, and he ended up only 9,000 miles away from his goal. This sure isn't India...

The simple fact is there are several issues with physical tokens and the way they are, and continue to be, implemented. Some of these problems have been around for almost the entire 30 years we've been using them.

Here are a few examples:

• Token deployment, even as simple as issuing a SecurID token to an employee, is tedious, and time consuming; sometimes even relying on snail mail which could take in excess of many months for distribution over an entire enterprise




• Over 10% to 15% will be damaged, lost, or stolen every year




• Barring all this, at best, a token's lifespan is 3 to 5 years




• Users forget their tokens - even ones attached to key rings; Contractors lose track of which token is for which client, adding more confusion




• Physical token systems require updates, maintenance, re-synchronization, and replacement

To every problem, a solution exists. How can we, as a global enterprise, still be secure, yet use technology that is already in place? The answer is simple - and you already own it.

Enter: The Mighty Cell Phone.

SMS has been in use for years, but with the global community reaching in excess of well over 5 billion devices currently in use, it seems most of us have overlooked one thing SMS is good for - that being to act as an authentication token. A passcode is sent to your Droid phone, for instance, thus eliminating the need for a physical token. SMS alone is not the answer, though. SMS does, in fact, have its pitfalls. The real answer lies in an ability to create similar (or even better, stronger) security that is already inherent in devices we already predominantly use. The unreliable nature of delays in messaging, cellular dead zones, or network issues with your provider are actual issues you may very well face. I am certain that those who wish to still cling to an antiquated technology, such as key fob based tokens, are sure to use as a rallying cry. At what point, though, do you stop bailing the sinking ship, and find a newer, better, stronger one?

Now Entering: MobilePASS.

MobilePASS by SafeNet, is a technology that still uses the same methodology of tokens, that being a two-factor authentication system. It simply does it in a better way. With MobilePASS, there's no additional hardware to lug around with you, the technology is easily deployed (and furthermore, managed), and the learning curve is non-existent. You do know how to download an app, don't you?

Strong encryption is still on the playing field. The tried and true two-factor authentication is still in the mix, yet the cumbersome, expensive, and dare I say, recently compromised hardware token, isn't.

"But what about my Windows Device?"

"There's an App for that".

In fact, not only is there an app for that old desktop of yours, but there's one for the iPhone, Blackberry, Windows Mobile, Android, Java 2ME, and even Out Of Band devices via SMS. Did I mention that it also covers Windows Server based platforms as well?

To log on to a secure network, from a laptop, PC or even an iPad (with the right software installed of course), users generate a One-Time Password (or OTP for short), via the MobilePASS app on their phone, and then enter this in the login screen thus creating a secure connection. Out of band delivery can also be granted via SMS or even email.

While MobilePASS by SafeNet is only one solution out there that utilizes cellular technology to end the need for the traditional token, this sort of solution is great for many reasons.

It is estimated that such a solution as moving to an SMS based method such as this can reduce the ongoing running costs of authentication means by over a whopping 40 to 60 percent! Lower cost, higher convenience, and utilizing technology already in place.

SafeNet is a global solutions partner of NWN Corporation. As such, NWN can leverage this partnership to assist our clients in lowering their bottom line, ensuring the security of their infrastructure, and delivering the best technology-based solutions currently available.

From ShmooCon - URL Enlargement

Well, ShmooCon started almost a week ago and it's about time to post something. First and foremost, all of the talks were outstanding and I'll probably wind up talking about most of them but I wanted to start with one that stuck out - URL Enlargement: Is it for You? by Daniel Crowley.

We all have seen URL shorteners. Many Twitter clients utilize them automatically. If you haven't seem one, check out www.tinyrul.com. Basically, you put a URL of any length in and you get a shortened version back out. For example, if I enter http://nwnsecurity.blogspot.com, I get back http://tinyurl.com/4v9e7w7. Now that's a savings of only 5 characters. That's not much but could help when using something like Twitter. If you are talking about longer URLs, the savings can be significant.

OK, that's convenient but what does that have to do with information security? Here's where the fun part begins. Check out this link - http://tinyurl.com/4g8gfpk No, really! I promise there's no malware there. In fact, it won't actually point to a web page. Just check out the URL that it resolves to. If you don't want to actually click on the like, check out http://www.longurlplease.com/ or http://www.longurl.org/.

OK, so we've established that URL shortening can be used to circumvent DLP systems. What else can it do? Well, a while back Billy Hoffman (a.k.a. Acidus) created a tool called TinyDisk. It is basically a file system that utilizes TinyURL (or other shortener) for storage. Now you can upload large data files to a stealth file system. That's cool. Unfortunately, I did some quick checking and couldn't find TinyDisk available for download. I'm sure it's out there but I couldn't find it with the 2 minutes of checking I did.

We've established that URL shortening can be used to establish covert channels but there are some other uses that I found to be particularly interesting.

When performing social engineering, will a user click on a link to http://www.somethingevil.com. Probably. But let's assume that the user is paying attention. They might think twice about clicking on somethingevil but what about clicking on http://tinyurl.com/o5h7wv. Now that's a completely different story. URL shorteners can be used to hide the javascript tags in a cross-site scripting attack or other URL parameters that might give away what you are really trying to do.

Also, when a user clicks on a shortened URL, their browser actually connects to the URL shortening service server and that server refers them to the final destination. If you are doing penetration testing and want to track who clicked on what, you can send different people URLs to the same site that were shortened by different URL shorteners. That way, tracking the referred will allow you to identify who clicked what.

One last point. What would happen if you encoded malware in base-64 and shortened that via a URL shortener? Hmmmmm.

Thanks Daniel for an excellent talk. Now off to the URL shortening.

Nessus Vulnerability XML Parser v8 and Cisco ACL Parser v0.05

It’s time for ShmooCon 2011, YEAH!!! This is my first time attending and I am very excited. I would like to release a few maintenance releases of my Nessus Vulnerability XML Parser v8 and Cisco ACL Parser v0.05.

Nessus Vulnerability XML Parser v8 – There was a bug in the creation of the TEXT File report generation. The issue was cause by a variable I called in a foreach loop, if the variable was not an array, but a hash the script would fail. No other changes were made.

The Cisco ACL Parser v0.05 – In the ASA a type of ACL used for the SSL Any-Connect Portal is called a WEBACL. There was a problem in the parsing of these ACL types. Also I changed the name of the output file to be the device hostname-output.csv “fw01-output.csv” instead of “hostname fw01-output.csv”.

These downloads can be found at Cody's blog at www.melcara.com.

I hope everyone enjoys the scripts and I hope to see you at ShmooCon.

Where did all the IP addresses go?

IANA owns the Internet. We'll, at least they manage who gets what IP addresses and guess what?!?!? We are running out. Check out Greg Ferro's posting at http://etherealmind.com/ipocalypse-what-next/ to get more of the scoop. The short story is that we are almost out of unallocated IP addresses requiring the move to IPv6. Now I don't have anything against IPv6 personally but I do know that a wholesale migration to a new IP addressing scheme will likely break one or two things and that's not something I'm looking forward to. That's why Greg's posting was interesting. He provided a link to an IANA web page that lists all of the IP address assignments (in terms of /8 networks). Curious, I checked it out.

I found a lot of what you'd expect - a bunch of /8 ranges assigned to ARIN, RIPE NCC, LACNIC, etc. That's not that interesting. I did find something that was, to me, more interesting. Over 20 /8 ranges were assigned to private companies such as GE, IBM, Xerox HP, DEC, Apple, MIT, Ford, Halliburton, Eli Lily, Bell North, Prudential Securities, Merck and duPont. Also, the Defense Information Systems Agency (DISA) has at least 4 that I counted. IANA has 2 allocated to themselves.

This got me thinking. Does any single company actually need over 16 million public IP addresses? I mean, in the age of NAT, PAT and RFC1918 what do you do with that many public networks. The 13 companies I listed above actually have over 200 million public IP addresses between them. To me, that's a little much.

Don't get me wrong. I'm not saying that these companies don't have a right to the IP addresses they registered. I'm not saying that these companies should be forced to give them up. I'm just wondering if DEC or Xerox even have 16 million computers, let alone the need for 16 million public addresses.

Wait! Stop! Hold on a minute! I'm thinking about this all wrong. Just think of the new security concerns there will be when we move to IPv6 in full. OK, so never mind what I was saying. The the IPv6 insanity begin!!!

Small Business Security 101 - Overview

Security is mandatory. This statement applies whether you are a fortune 50 global mega-corporation or a small business with 25 employees. As long as you rely on computers and connect to the Internet, you face computer-related risks. Actually, even if you don't connect to the Internet, those risks are present. Infection by a virus, worm or spyware, compromise by an external attacker, malicious insiders, careless employees and even hardware failures are threats that organization of all sizes must face. Large organizations have resources to deal with these threats but small and medium-sized business don't often that the money, technology, skills or experience necessary to affect reasonable levels of security. This is the first in a series of posting designed to provide an overview of how organizations with limited time, limited budget and limited personnel can achieve necessary security and, as necessary, regulatory compliance. The focus will be on "cost effective security" that introduces a minimum of operational overhead at a minimum of cost.

This first posting is going to address some of the myths about security and small to medium-sized businesses so let's get right down to business.


Myth 1: My business is too small to be targeted.

This often (but not exclusively) comes up when discussing security with organizations that have only a few employees. The thought is that "hackers" will target big companies and will ignore small ones. This seems to make sense. Big companies have more to steal, they have a larger footprint and they are far more well known. This, however, assumes that all attackers are intentionally targeting their victims - a fact that is not always true. There are plenty of "target of opportunity" attackers out there. Viruses and worms, for the most part, don't care who they infect and if an attacker finds your vulnerable system during random scanning of the Internet, they will be happy to compromise you. Making things worse, your users are browsing the Internet, taking advantage of wireless hot spots at airports, hotels and Starbucks and placing their laptops behind Best Buy purchased broadband routers with the default configuration when they work from home. All of these put your users and your organization at risk. Some studies have shown that a random computer gets compromised if placed unprotected on the Internet after 15 minutes. There are 96, 15-minute blocks per day, 672 per week and 34,944 per year. That's a lot of time for a compromise to happen.

Myth 2: I don't have anything of value so there's no reason I'd be compromised.

This statement simply highlights some misunderstandings or ignorance about what is considered valuable. To a bad buy hacker, any business is ripe with valuable targets. They want your processing power, Internet bandwidth and hard drive space for storing and distributing porn, stolen credit card numbers and other contraband. They also want your systems to use as Spam relays and bot-net zombies.

More personally, employees have access to bank accounts (corporate and personal), Facebook accounts and and even World of Warcraft. Recently, a "friend" on Facebook contacted me via Facebook IM stating that he got mugged outside his hotel in London and that he needed money to get home. His account had been compromised. I didn't send any money but how many people would have - particularly if it was a close friend. I'll also admit to the fact that I had a World of Warcraft account (yes I'm a nerd) that I stopped using for around a year or so get compromised. These things happen all the time.

Finally, bad guy hackers also want personal information such as names, addresses and social security numbers. Identity theft is real and can create havoc in the lives of the victims. And if you think that one of your employees suffering identity theft has nothing to do with you business, wait until it happens and see how their performance and productivity are affected.

Myth3: I have a firewall and anti-virus so I'm secure.

This myth was not limited to the SMB space until fairly recently. This was a common belief until significant, publicized breaches and comprehensive regulatory requirements swayed many large (and regulated) organizations over to a more comprehensive security program mentality. Unfortunately, it still rings true for many organizations. It is, in fact, quite false for a number of reasons. First, the fact is that anti-virus technology is not perfect. Every time testing is done, testers find that some AV products catch some malware and some miss. I've not seen any solid statistics but the number that rings fairly true is that the average AV product is between 60 and 70% effective. Anti-virus is not the answer (although it is part of the answer).

Firewalls are even more of a problem. In the past firewalls were set up to block all inbound traffic and that was good. Then organizations decided it would be a good thing to receive email and host web sites - and holes were created in the firewall. Then organizations went from web sites to mission critical web applications and more holes were created. Today, it is not uncommon to find firewalls with dozens or ports and protocols allowed in, each and every one of which represent a risk. Making things worse, most firewalls are configured to block (at least some) inbound access but to allow all outbound access. This means that users all allowed to establish outbound connections to any IP address on the Internet using any port - and the firewall will allow the responses back in. All an attacker need to do is trick one user into opening a file, click on a link or visit a malicious site and your firewall is now, effectively useless.

Myth 4: Small businesses have simple IT environments.

This isn't something that people often say but rather something that people often assume. Along with that assumption is that security is, as a result, easy. While this may be the case for some organizations, it is not uncommon for even fairly small organizations to have VoIP phone systems, virtualization, wireless, storage (either SAN or NAS), wide area network connections, VPN (site-to-site, IPSec and SSL) tunnels and Active Directory (or other LDAP). These are complex and sophisticated technologies that each bring with them a unique set of security concerns. Particularly in smaller organizations, knowledge of these security risks and the skills necessary to address them is not available making security all the more difficult.

Summary
We've discussed some of the myths about security for small to medium-sized businesses. Hopefully, it is fairly obvious that in the SMB space, while organizations may have fewer employees, they require every bit as sophisticated a security program as their larger counterparts. Consider the following:

• A collections company that maintains a database with the personal information of over 11 million people including, in some cases, credit card and bank account data
• A private equity fund company that manages over $3 billion in assets
• A web application development company that focuses on providing health care data to pharmaceutical companies
• A business that provides data center services to credit unions

All of these are real business that have two things in common. They all have sophisticated security needs and they all have less than 5o employees.

The question faced by these organzations, and by may others is how can we achieve these security goals with minimal staff, minimal budget and minimal resources. That is the question that will be answered by upcoming postings. In future postings we will discuss each of a variety of security topics and how they can be addressed by the average SMB. We'll be discussing the following:

• VoIP – network design, 802.1x & NAC
• IDS/IPS
• Endpoint - anti-malware, encryption, application control (blacklisting/whitelisting), integrity verification
• Smart phones
• Network design
• Web app vulnerability scanning, web app firewall
• Firewalls and VPN
• Authentication – user provisioning, good passwords, tokens, rights assignment, admin rights
• SEIM/Central Monitoring
• Network Device Hardening
• Active Directory – design, GPO
• Vulnerability Scanning, Patch Management
• Virtualization
• Wireless network design, authentication/802.1x/WPA2, etc.
• Wireless client attacks, bluetooth, keyboards, etc.
• Wireless IDS
• Data Leakage Protection
• Media Sanitization
• Data Classification, Risk Assessment, Security Awareness, Acceptable Use, Incident Response, Change Control
• Physical Security
• Regulatory Compliance, Policies

For each, we will discuss a typical SMB environment, the risks, the controls and how best to accomplish cost-effective security. Hope to see you there.

Auditing Access Control Lists

During assessments NWN regularly reviews the configurations of many different types of systems and devices ranging from firewalls, routers, switches and servers. Many of these devices, particularly routers, switches and firewalls, have Access Control Lists or ACLs that provide critical security controls. Unfortunately, the process of auditing these ACLs can sometimes be very time consuming. Until now there has not been an open source or free tool to assist with this type of auditing. Fortunately, member of the NWN STAR team, Cody Dumont, has created a tool - the ACL2CSV parser.

ACL2CSV parser is a PERL script, which reads Cisco router, switch, and firewall (PIX, and ASA) configuration files that have been exported from the device in flat text format. It parses the ACLS and object groups (PIX and ASA only) and generates an easy to understand CSV file. This file can them be opened in Microsoft Excel or other spreadsheet software for easy viewing or additional analysis. The ACL2CSV tool expands all object groups and places them into the correct location as if the ACL did not use object groups. The object group expansion works for all object group types. The ACL2CSV parser is extremely fast and easy to use.

For instructions on the use of ACL2CSV, visit Cody’s personal blog at http://www.melcara.com and select the link for Cisco ACL2CSV parser. If PERL is not installed on your system, you will need to do so. PERL can be found at http://www.activestate.com or http://www.perl.com. Once PERL is installed no additional modules are needed. The script assumes PERL is installed at “/usr/bin/perl”, and only uses the “strict” and “Getopt::Std” modules. After the script is copied into a folder found in your system path, you should also modify the permissions to allow the script to be executable. On a Unix, Linux or Mac computer, this can be accomplished via the chmod command.

Once the script is executed and installed, you can run the script. The script requires a file name as an argument, so if you simply type ./acl2csv.pl you will receive the following message:

The command requires a File Name as a command line argument
acl2csv.pl c:\old_pix_config.txt

To check the version, add the “-v” or “v” as the command line argument.
ACL Parser for Cisco IOS, PIX & ASA
DEVELOPED AND OWNED BY Cody Dumont - NWN Security Testing Assessment and Response (STAR)
Licensed to Planet Earth.
I used some code for another tool of this type from James Bly AT mangeek.com
http://mangeek.com/portfolio/pixparser.html
Also Anthony contributed by doing some testing and verification
Version Number 0.04 - Dec 2010
"For Questions Please Contact Cody Dumont - CDumont@nwnit.com "


To run the script, use the following command:
/foo/bar/acl2csv.pl fw_config.txt


Once the script is finished, a completion message will be displayed. There should be no other messages. In the event you receive a PERL array or HASH error message, this means that Cody did not do enough testing. If you get such an error, follow the PERL debugging steps or contact Cody at CDumont@nwnit.com and he’ll be happy to fix the error.
The output of the script will look something like the following.

NAME,LINE,TYPE,FUNCTION,PROTOCOL,SOURCE NET,SOURCE_PORT,DEST NET,DEST PORT,TIME,INACTIVE,LOG,REMARK,ORIGINAL
test441,1,extended,permit,tcp,any,,any,eq 44 ,,,,,access-list test441 extended permit tcp any any eq 44
test442,1,extended,permit,tcp,any,eq 44,any,eq 44 ,,,,,access-list test442 extended permit tcp any eq 44 any eq 44
test443,1,extended,permit,tcp,any,eq 44,any, ,,,,,access-list test443 extended permit tcp any eq 44 any

The definitions of each field are listed below.
NAME – The name of the ACL
LINE – The line in sequential order of the ACL. Please note if the ACL uses Object-Groups, then each Object-Group will have the same index number. The idea is to provide the user with same output of the “show access-lists” command. Please note “REMARK” ACL entries are not counted in this test.
TYPE – This is the TYPE of ACL, the options are standard, extended, webtype.
FUNCTION – The “permit” or “deny” entry.
PROTOCOL – The Layer 3 protocol controlled by the ACL entry.
SOURCE NET – The source network and subnet mask. If the entry is a host, then “host” will be displayed. However if “255.255.255.255” is found in the ACL entry, then “255.255.255.255” will be displayed. The script does not check the validly of the mask, and the assumption is the config is a direct output from “show running-configuration” or “show startup-configuration”.
SOURCE_PORT – If a source port is defined, then the source port is displayed. If no source port is defined this field will be empty.
DEST NET – Same as the SOURCE NET, but the destination section of the ACL entry is displayed.
DEST PORT - Same as the SOURCE PORT part of the ACL, but the destination section of the ACL entry is displayed.
TIME – ACLs can be time sensitive using the “time-range” command. If a time-range is defined, the name of the “time-range” is displayed, however the details of the “time-range” are not displayed. This might be added in a future release.
INACTIVE – If an ACL entry is listed as “INACTIVE”, the entry is left in the configuration, but is not an active rule. Other parsing tools often ignore this, but if “YES” is found in this field then the ACL enter is not an active rule. If the field is empty then the ACL entry is an active rule.
LOG – If log settings are configured, the settings are displayed in this cell. Two examples are “interval 5” and “notifications”.
REMARK – The “REMARK” section is a little harder to deal with. The configuration files currently put the “REMARK” in an ACL entry just above the ACL entry the “REMARK” is connected to. So in the parsing of the ACL, the script will check to see if the preceding entry was a remark, if so this field will be filled with “REMARK” statement. However some ACL’s might have more than one line as a remark, the script will not detect this case. The script will only detect the “REMARK” in the preceding line only.
ORIGINAL – The original ACL entry unmodified. The “REMARK” ACL entries are not displayed.

When the script is finished, open the file using any spreadsheet application. Then you can create filters, freeze panes, etc. You could also import the newly created ACL file into a database.

This tool can save a network administrator, security professional or auditor a lot of time sorting through ACLs. We at NWN STAR hope the security community will find this tool useful and will enhance the overall security of your information system.