NETRESEC Network Security Blog - Tag : domain

rss Google News

Reassembling Victim Domain Fragments from SUNBURST DNS

We are releasing a free tool called SunburstDomainDecoder today, which is created in order to help CERT organizations identify victims of the trojanized SolarWinds software update, known as SUNBURST or Solorigate.

SunburstDomainDecoder.exe output showing innout.corp nswhealth.net cisco.com fa.lcl int.lukoil-international.uz tr.technion.ac.il bisco.int phabahamas.org banccentral.com bk.local htwanmgmt.local

SunburstDomainDecoder can be fed with DNS queries to avsvmcloud.com in order to reveal the full internal domain names of infected companies and organizations.

UPDATE December 18, 2020 (v1.1)

SunburstDomainDecoder has now been updated to automatically reassemble fragmented domain name segments in order to show the full domain in the output.

UPDATE December 19, 2020 (v1.2)

Domain names that have been base32 encoded, such as domain names with uppercase letters, can now be extracted with SunburstDomainDecoder. The queried SUNBURST subdomains are now also included in the output.

UPDATE December 21, 2020 (v1.6)

Improved parsing of base32 encoded domain names. SUNBURST victim domains like "LKDataCenter.com", "Sunkistgrowers.com" and "BrokenArrow.Local" can now be extracted.

UPDATE December 27, 2020 (v1.7)

Improved reassembly of long domain names, like "CIMBMY.CIMBDomain.com" and "BE.AJINOMOTO-OMNICHEM.AD", that get segmented into multiple parts. Extraction of time stamps and security applications, including "Windows Defender", "Carbon Black", "CrowdStrike", "FireEye", "ESET" and "F-Secure". See Sergei Shevchenko's blog post Sunburst Backdoor, Part III: DGA & Security Software for more details.

UPDATE January 4, 2021 (v1.8)

Security products (WinDefend, ESET etc.) are now included in the summary output at the end. SUNBURST stage2 victims, which accept C2 domains in CNAME responses, are indicated with a "STAGE2" tag. The previous release marked stage2 queries with a "DNSSEC" tag. Improved extraction of truncated base32 domains, such as "*TED.com".

UPDATE January 12, 2021 (v1.9)

DNS queries with encoded timestamps are tagged with either "AVProducts" or "Ping", depending on if they include an update of the installed/running security products and services or not. The summary data at the end has been modified to also show partial domain names, such as "paloaltonetworks*".

UPDATE February 16, 2021 (v2.0)

Slightly faster and even more accurate than previous versions.

Download SunburstDomainDecoder.zip

 

SUNBURST DNS Traffic

SUNBURST victims, who have installed one of the trojanized SolarWinds Orion software updates, will query for domain names formatted like this:

<SUBDOMAIN>.appsync-api.eu-west-1.avsvmcloud.com
<SUBDOMAIN>.appsync-api.us-west-2.avsvmcloud.com
<SUBDOMAIN>.appsync-api.us-east-1.avsvmcloud.com
<SUBDOMAIN>.appsync-api.us-east-2.avsvmcloud.com

The "SUBDOMAIN" string has different values for each victim and the second half of this string actually contains an encoded domain name (encrypted with a simple substitution cipher).

RedDrip's decode.py

The RedDrip Team published a SUNBURST DGA decoding script yesterday, which can be used to identify SUNBURST victim organizations like CISCO and Belkin by decoding the domain names encoded in the outgoing DNS queries for subdomains of avsvmcloud.com.

This is what it looks like when RedDrip's decode.py script is fed with domain names from John Bambenek's uniq-hostnames.txt file.

cat uniq-hostnames.txt | python decode.py
02m6hcopd17p6h450gt3.appsync-api.us-west-2.avsvmcloud.com .gh
039n5tnndkhrfn5cun0y0sz02hij0b12.appsync-api.us-west-2.avsvmcloud.com ad001.mtk.lo
04spiistorug1jq5o6o0.appsync-api.us-west-2.avsvmcloud.com isi
060mpkprgdk087ebcr1jov0te2h.appsync-api.us-east-1.avsvmcloud.com belkin.com
06o0865eliou4t0btvef0b12eu1.appsync-api.us-east-1.avsvmcloud.com gncu.local
07605jn8l36uranbtvef0b12eu1.appsync-api.us-east-1.avsvmcloud.com gncu.local
07q2aghbohp4bncce6vi0odsovertr2s.appsync-api.us-east-1.avsvmcloud.com csnt.princegeor
07ttndaugjrj4pcbtvef0b12eu1.appsync-api.us-east-1.avsvmcloud.com gncu.local
08amtsejd02kobtb6h07ts2fd0b12eu1.appsync-api.eu-west-1.avsvmcloud.com sm-group.local
0b0fbhp20mdsv4scwo11r0oirssrc2vv.appsync-api.us-east-2.avsvmcloud.com ville.terrebonn
[...]

The beauty of this approach is that passive DNS data can be used in order to reliably identify the victims. This is great news for national CERTs, because they typically have readily access to passive DNS data and can use the decoded domain names in order to identify and reach out to victims in their country.

After using the python script provided by ReadDrip Team I noticed two things:

  1. The leaked domain names were internal domain names used on the victim organizations' corporate networks. Many of the domains were using the ".local" suffix.
  2. Most of the extracted domains were truncated to around 15 bytes, which make it difficult to identify the victim organization.

Truncated Domains Fragmented Domains

I later learned that what seemed to be truncated domains were actually fragmented domains, where long domain names would be split into multiple queries. This revelation turns the output from RedDrip's python tool into an interesting domain name puzzle. At this point I decided to take a closer look at the malicious SolarWinds update I had downloaded from SolarWind's website a few days ago -- yes, that's right the malicious software update "SolarWinds-Core-v2019.4.5220-Hotfix5.msp" (MD5: 02af7cec58b9a5da1c542b5a32151ba1) was actually available for download from SolarWinds' website long after they had been notified about their software being backdoored!

As an example, lets' take a closer look at this DNS query from John Bambenek's passive DNS data:
r1qshoj05ji05ac6eoip02jovt6i2v0c.appsync-api.us-west-2.avsvmcloud.com

This query can be broken down into three parts:

  1. r1qshoj05ji05ac6 : What is encoded here???
  2. eoip02jovt6i2v0c : Base32 encoded string "city.kingston."
  3. .appsync-api.us-west-2.avsvmcloud.com : DNS trailer without encoded data

So, which "City of Kingston", or "Kingston City", should we contact to let them know that they have installed a trojanized SolarWinds update? Is it Kingston Jamaica, City of Kingston NY USA, City of Kingston Ontario Canada, Kingston City Tennessee USA or City of Kingston Australia?

After analyzing the "SolarWinds.Orion.Core.BusinessLayer.dll" file (MD5: b91ce2fa41029f6955bff20079468448) from the "SolarWinds-Core-v2019.4.5220-Hotfix5.msp" I learned that the initial "r1qshoj05ji05ac6" string is representing a unique "GUID" value for the infected machine. This GUID is generated by calculating an MD5 hash of the MAC address of the first active non-Loopback network interface, the domain name and the "MachineGuid" registry key value in "HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Cryptography".

This MD5 hash is then squeezed into a tiny 8 byte array by XOR'ing overlapping bytes. The "CreateSecureString" function in the trojanized SolarWinds update then "encrypts" this hash using XOR with a random key, which is prepended to the data. The XOR key and the XOR'ed data is then finally base32 encoded into what makes up the first part of the subdomain to query for. Don't let the SUNBURST source code below fool you, it is actually using base32 encoding with a custom alphabet even though the function is called "Base64Encode";

CreateSecureString function in SolarWinds.Orion.Core.BusinessLayer.OrionImprovementBusinessLayer.CryptoHelper
Image: SUNBURST source code generates a random value between 1 and 127 as XOR key

Each DNS lookup from an infected machine will query for a unique subdomain because a new XOR key will be generated for each request. Luckily for us, this XOR key is provided in each request, so we can use it in order to "decrypt" the subdomain and get the original 8 bytes derived from the MAC+domain+MachineGuid MD5 hash.

The output from my "SunburstDomainDecoder.exe" tool will print the "decrypted" 8 byte GUID in the first column, the decoded victim domain segment or timestamp in the second column and the queried SUNBURST subdomain in the last column. Each DNS query line read from standard input will generate a "GUID DecodedHostname SunburstSubdomain" line on standard output.

SunburstDomainDecoder.exe < uniq-hostnames.txt
F18613981DEC4D1A 2020-10-02T21:00:00.0000000Z 02m6hcopd17p6h450gt3
BD6DEFBBE9FEA3A9 ad001.mtk.lo 039n5tnndkhrfn5cun0y0sz02hij0b12
2BF8DE15406EA780 2020-08-25T03:00:00.0000000Z 043o9vacvthf0v95t81l
573DEB889FC54130 2020-08-13T21:00:00.0000000Z,​WindowsDefender_RUNNING,CrowdStrike_RUNNING 04jrge684mgk4eq8m8adfg7
518092C8FD571806 2020-06-09T22:30:00.0000000Z 04r0rndp6aom5fq5g6p1
F18613981DEC4D1A 2020-07-06T08:30:00.0000000Z 04spiistorug1jq5o6o0
BC1CB013239B4B92 2020-04-25T10:00:00.0000000Z 05q2sp0v4b5ramdf71l7
3ED2E979D53B2523 belkin.com 060mpkprgdk087ebcr1jov0te2h
4225A5C345C1FC8E gncu.local 06o0865eliou4t0btvef0b12eu1
[...]

The tool then finishes off by outputting the domains that are complete or at least have the last part of their domain intact. Some of these domains are complete because they were short enough to fit in one single SUNBURST DNS query, while others have been pieced together by SunburstDomainDecoder from domain fragments arriving in separate SUNBURST DNS queries.

[...]
F59BBAACBA3493C0 dufferincounty.on.ca
F5D6AA262381B084 glu.com
F9024D5B1E9717C6 gyldendal.local
F90BDDB47E495629 central.pima.gov
F956B5EF56BCF666 coxnet.cox.com
F9A9387F7D252842 city.kingston.on.ca
FB0B50553BC00DED gloucesterva.net
FBB6164BC2B0DFAD ARYZTA.COM
FD04AC52C95A1B0A bmrn.com
FDFCAB8E4C0AB3EE ansc.gob.pe
FE7FF8C9104A0508 thoughtspot.int
FF6760F36DB3D7DC smes.org

We can now see that it was "city.kingston.on.ca", (City of Kingston, Ontario, Canada) who had installed a trojanized SolarWinds update.

Download SunburstDomainDecoder

The C# source code and a compiled Windows binary for SunburstDomainDecoder is available here:
https://www.netresec.com/files/SunburstDomainDecoder.zip

Creative Commons CC-BY

The source code and Windows binary is shared under a Creative Commons CC-BY license, which means that you are free to:

  • Share : copy and redistribute the material in any medium or format
  • Adapt : remix, transform, and build upon the material for any purpose, even commercially.
Provided that you give appropriate credit, provide a link to the license, and indicate if changes were made.

Running SunburstDomainDecoder on Linux/MacOS

Wanna run SunburstDomainDecoder.exe but not in Windows? No problems, the tool runs perfectly fine in Mono. Another option is to build SunburstDomainDecoder.cs as a .NET core project in Linux.

.NET Reversing

Would you like to verify my findings or learn more about .NET reverse engineering? Cool, then I'd recommend that you download dnSpy in order to reverse engineer the SUNBURST .NET DLL (which can be extracted from the msp installer with 7zip). Or you can have a look at the already extracted OrionImprovementBusinessLayer.cs on GitHub.

Posted by Erik Hjelmvik on Thursday, 17 December 2020 22:30:00 (UTC/GMT)

Tags: #SunburstDomainDecoder#SUNBURST#SolarWinds#Solorigate#domain#DNS#pDNS#Windows Defender#Carbon Black#FireEye#ESET#F-Secure#Trojan#avsvmcloud

Short URL: https://netresec.com/?b=20C0f71


Domain Whitelist Benchmark: Alexa vs Umbrella

Alexa vs Umbrella

In November last year Alexa admitted in a tweet that they had stopped releasing their CSV file with the one million most popular domains.

Yes, the top 1m sites file has been retired

Members of the Internet measurement and infosec research communities were outraged, surprised and disappointed since this domain list had become the de-facto tool for evaluating the popularity of a domain. As a result of this Cisco Umbrella (previously OpenDNS) released a free top 1 million list of their own in December the same year. However, by then Alexa had already announced that their “top-1m.csv” file was back up again.

The file is back for now. We'll post an update before it changes again.

The Alexa list was unavailable for just about a week but this was enough for many researchers, developers and security professionals to make the move to alternative lists, such as the one from Umbrella. This move was perhaps fueled by Alexa saying that the “file is back for now”, which hints that they might decide to remove it again later on.

We’ve been leveraging the Alexa list for quite some time in NetworkMiner and CapLoader in order to do DNS whitelisting, for example when doing threat hunting with Rinse-Repeat. But we haven’t made the move from Alexa to Umbrella, at least not yet.


Malware Domains in the Top 1 Million Lists

Threat hunting expert Veronica Valeros recently pointed out that there are a great deal of malicious domains in the Alexa top one million list.

Researchers using Alexa top 1M as legit, you may want to think twice about that. You'd be surprised how many malicious domains end there.

I often recommend analysts to use the Alexa list as a whitelist to remove “normal” web surfing from their PCAP dataset when doing threat hunting or network forensics. And, as previously mentioned, both NetworkMiner and CapLoader make use of the Alexa list in order to simplify domain whitelisting. I therefore decided to evaluate just how many malicious domains there are in the Alexa and Umbrella lists.

hpHosts EMD (by Malwarebytes)

Alexa Umbrella
Whitelisted malicious domains: 1365 1458
Percent of malicious domains whitelisted: 0.89% 0.95%

Malware Domain Blocklist

Alexa Umbrella
Whitelisted malicious domains: 84 63
Percent of malicious domains whitelisted: 0.46% 0.34%

CyberCrime Tracker

Alexa Umbrella
Whitelisted malicious domains: 15 10
Percent of malicious domains whitelisted: 0.19% 0.13%

The results presented above indicate that Alexa and Umbrella both contain roughly the same number of malicious domains. The percentages also reveal that using Alexa or Umbrella as a whitelist, i.e. ignore all traffic to the top one million domains, might result in ignoring up to 1% of the traffic going to malicious domains. I guess this is an acceptable number of false negatives since techniques like Rinse-Repeat Intrusion Detection isn’t intended to replace traditional intrusion detection systems, instead it is meant to be use as a complement in order to hunt down the intrusions that your IDS failed to detect. Working on a reduced dataset containing 99% of the malicious traffic is an acceptable price to pay for having removed all the “normal” traffic going to the one million most popular domains.


Sub Domains

One significat difference between the two lists is that the Umbrella list contains subdomains (such as www.google.com, safebrowsing.google.com and accounts.google.com) while the Alexa list only contains main domains (like “google.com”). In fact, the Umbrella list contains over 1800 subdomains for google.com alone! This means that the Umbrella list in practice contains fewer main domains compared to the one million main domains in the Alexa list. We estimate that roughly half of the domains in the Umbrella list are redundant if you only are interested in main domains. However, having sub domains can be an asset if you need to match the full domain name rather than just the main domain name.


Data Sources used to Compile the Lists

The Alexa Extension for Firefox
Image: The Alexa Extension for Firefox

The two lists are compiled in different ways, which can be important to be aware of depending on what type of traffic you are analyzing. Alexa primarily receives web browsing data from users who have installed one of Alexa’s many browser extensions (such as the Alexa browser toolbar shown above). They also gather additional data from users visiting web sites that include Alexa’s tracker script.

Cisco Umbrella, on the other hand, compile their data from “the actual world-wide usage of domains by Umbrella global network users”. We’re guessing this means building statistics from DNS queries sent through the OpenDNS service that was recently acquired by Cisco.

This means that the Alexa list might be better suited if you are only analyzing HTTP traffic from web browsers, while the Umbrella list probably is the best choice if you are analyzing non-HTTP traffic or HTTP traffic that isn’t generated by browsers (for example HTTP API communication).


Other Quirks

As noted by Greg Ferro, the Umbrella list contains test domains like “www.example.com”. These domains are not present in the Alexa list.

We have also noticed that the Umbrella list contains several domains with non-authorized gTLDs, such as “.home”, “.mail” and “.corp”. The Alexa list, on the other hand, only seem to contain real domain names.


Resources and Raw Data

Both the Alexa and Cisco Umbrella top one million lists are CSV files named “top-1m.csv”. The CSV files can be downloaded from these URL’s:

The analysis results presented in this blog post are based on top-1m.csv files downloaded from Alexa and Umbrella on March 31, 2017. The malware domain lists were also downloaded from the three respective sources on that same day.

We have decided to share the “false negatives” (malware domains that were present in the Alexa and Umbrella lists) for transparency. You can download the lists with all false negatives from here:
https://www.netresec.com/files/alexa-umbrella-malware-domains_170331.zip


Hands-on Practice and Training

If you wanna learn more about how a list of common domains can be used to hunt down intrusions in your network, then please register for one of our network forensic trainings. The next training will be a pre-conference training at 44CON in London.

Posted by Erik Hjelmvik on Monday, 03 April 2017 14:47:00 (UTC/GMT)

Tags: #Alexa#Umbrella#domain#Threat Hunting#DNS#malware

Short URL: https://netresec.com/?b=1743fae

X / twitter

X / Twitter: @netresec


Bluesky

Bluesky: @netresec.com


Mastodon

Mastodon: @netresec@infosec.exchange