rm -rf $ALLTHETHINGS/ | TechSNAP 262

chris

6 years ago

Find out why everyone’s just a little disappointed in Badlock, the bad security that could be connected to the Panama Papers leak & the story of a simple delete command that took out an entire hosting provider.

Plus your batch of networking questions, our answers & a packed round up!

Thanks to:

Get Paid to Write for DigitalOcean

Direct Download:

RSS Feeds:

Become a supporter on Patreon:

Show Notes:

Badlock vulnerability disclosed

The badlock vulnerability was finally disclosed on Tuesday after 3 weeks of hype
It turns out to not have been as big a deal as we were lead to believe
The flaw was not in the SMB protocol itself, but in the related SAM and LSAD protocols
The flaw itself is identified as https://web.nvd.nist.gov/view/vuln/detail?vulnId=CVE-2016-2118
It affects all versions of Samba clear back to 3.0
“Samba 4.4.2, 4.3.8 and 4.2.11 Security Releases are available”
“Please be aware that Samba 4.1 and below are therefore out of support, even for security fixes. There will be no official security releases for Samba 4.1 and below published by the Samba Team or SerNet (for EnterpriseSAMBA). We strongly advise users to upgrade to a supported release.”
See the Samba Release Planning page for more details about support lifetime for each branch
Microsoft releases MS16-047 but rated it only “Important”, not “Critical”
The patch fixes an “elevation of privilege bug in both SAM and LSAD that could be exploited in a man-in-the-middle attack, forcing a downgrade of the authentication level of both channels. An attacker could then impersonate an authenticated user”
Microsoft was also careful to note: “Only applications and products that use the SAM or LSAD remote protocols are affected by this issue. The SMB protocol is not vulnerable.”
It seems most of the “badlock” bugs were actually in Samba itself, rather than the protocol as we were lead to believe
“There are several MITM attacks that can be performed against a variety of protocols used by Samba. These would permit execution of arbitrary Samba network calls using the context of the intercepted user. Impact examples of intercepting administrator network traffic:”
Samba AD server – view or modify secrets within an AD database, including user password hashes, or shutdown critical services.
standard Samba server – modify user permissions on files or directories.
There were also a number of related CVEs that are also fixed:
- CVE-2015-5370 3.6.0 to 4.4.0: Errors in Samba DCE-RPC code can lead to denial of service (crashes and high cpu consumption) and man in the middle attacks. It is unlikely but not impossible to trigger remote code execution, which may result in an impersonation on the client side.
- CVE-2016-2110 3.0.0 to 4.4.0: The feature negotiation of NTLMSSP is not downgrade protected. A man in the middle is able to clear even required flags, especially NTLMSSP_NEGOTIATE_SIGN and NTLMSSP_NEGOTIATE_SEAL. Which has implications on encrypted LDAP traffic.
- CVE-2016-2111 3.0.0 to 4.4.0: When Samba is configured as Domain Controller it allows remote attackers to spoof the computer name of a secure channel’s endpoints, and obtain sensitive session information, by running a crafted application and leveraging the ability to sniff network traffic.
- CVE-2016-2112 3.0.0 to 4.4.0: A man in the middle is able to downgrade LDAP connections to no integrity protection. It’s possible to attack client and server with this.
- CVE-2016-2113 4.0.0 to 4.4.0: Man in the middle attacks are possible for client triggered LDAP connections (with ldaps://) and ncacn_http connections (with https://).
- CVE-2016-2114 4.0.0 to 4.4.0: Due to a bug Samba doesn’t enforce required smb signing, even if explicitly configured. In addition the default for the active directory domain controller case was wrong.
- CVE-2016-2115 3.0.0 to 4.4.0: The protection of DCERPC communication over ncacn_np (which is the default for most the file server related protocols) is inherited from the underlying SMB connection. Samba doesn’t enforce SMB signing for this kind of SMB connections by default, which makes man in the middle attacks possible.
Additional Coverage: Threadpost – Badlock vulnerability falls flat against its type
“As it turns out, Badlock was hardly the remote code execution monster many anticipated. Instead, it’s a man-in-the-middle and denial-of-service bug, allowing an attacker to elevate privileges or crash a Windows machine running Samba services.”
“Red Hat security strategist Josh Bressers said Badlock could have been much worse, especially if it had turned out to be a memory corruption issue in SMB as some had surmised. Such a scenario would have cleared a path for remote code execution, for example.”
Additional Coverage: sadlock.org

Panama Papers: Mossack Fonseca

Eleven million documents were leaked from one of the world’s most secretive companies, Panamanian law firm Mossack Fonseca.
They show how Mossack Fonseca has helped clients launder money, dodge sanctions and avoid tax.
The documents show 12 current or former heads of state and at least 60 people linked to current or former world leaders in the data.
Eleven million documents held by the Panama-based law firm Mossack Fonseca have been passed to German newspaper Sueddeutsche Zeitung, which then shared them with the International Consortium of Investigative Journalists. BBC Panorama is among 107 media organisations – including UK newspaper the Guardian – in 76 countries which have been analysing the documents.
There are many conspiracy theories about the source of the Panama Papers leak. One of the more prominent theories today blames the CIA.
Bradley Birkenfeld is “the most significant financial whistleblower of all time,” and he has opinions about who’s responsible for leaking the Panama Papers rattling financial and political power centers around the world.
Wikileaks is also getting attention today for blaming USAID and George Soros for the leaks.
What little is known about the source of the leak comes from details published by German newspaper Suddeutsche Zeitung. Communicating via encrypted chat in late 2014, the source warned his or her life was “in danger” but that they had data from law firm Mossack Fonseca that they wanted to share. When asked how much data they had, the source replied “more than you have ever seen,” according to the newspaper.
Regardless, the front-end computer systems of Mossack Fonseca are outdated and riddled with security flaws, analysis has revealed.
Mossack Fonseca’s client portal is also vulnerable to the DROWN attack, a security exploit that targets servers supporting the obsolete and insecure SSL v2 protocol. The portal, which runs on the Drupal open source CMS, was last updated in August 2013, according to the site’s changelog.
On its main website Mossack Fonseca claims its Client Information Portal provides a “secure online account” allowing customers to access “corporate information anywhere and everywhere”. The version of Drupal used by the portal has at least 25 vulnerabilities, including a high-risk SQL injection vulnerability that allows anyone to remotely execute arbitrary commands. Areas of the portal’s backend can also be accessed by guessing the URL structure, a security researcher noted.
Mossack Fonseca’s webmail system, which runs on Microsoft’s Outlook Web Access, was last updated in 2009, while its main site runs a version of WordPress that is three months out of date. A further vulnerability makes it possible to easily access files uploaded to the backend of Mossack Fonseca’s site simply by guessing the URL.
Mossack Fonseca’s emails were also not transport encrypted, according to privacy expert Christopher Soghoian who noted the company did not use the TLS security protocol.
Who leaked the Panama Papers? A famous financial whistleblower says: CIA. / Boing Boing
Wikileaks Accuses US Of Funding Panama Papers Putin Expose | The Daily Caller
Panama Papers: The security flaws at the heart of Mossack Fonseca (Wired UK)
Additional Coverage: The Register – Mossack Fonseca website found vulnerable to SQL injection
Additional Coverage: Forbes
Additional Coverage: WordFence
Additional Coverage: Slashdot
In general, it seems there were so many flaws in the website we may never know which one was used to compromise the server

I accidently rm -rf /’d, and destroyed my entire company

“I run a small hosting provider with more or less 1535 customers and I use Ansible to automate some operations to be run on all servers. Last night I accidentally ran, on all servers, a Bash script with a rm -rf {foo}/{bar} with those variables undefined due to a bug in the code above this line.”
“All servers got deleted and the offsite backups too because the remote storage was mounted just before by the same script (that is a backup maintenance script).
How I can recover from a rm -rf / now in a timely manner?”
There is not usually any easy way to recover from something like this
That is why you need backups. Backups are not just a single copy of your files in another location, you need time series data, in case you need to go back more than the most recent backup
It is usually best to not have your backups mounted directly, for exactly this reason
Even if you will never rm -rf /, an attacker might run rm -rf /backup/*
While cleaning up after an attacker attempted to use a Linux kernel exploit against my FreeBSD machine in 2003, I accidently rm -rf /’d in a roundabout way, Trying to remove a symlink to / that had a very funky name (part of the exploit iirc), i used tab complete, and instead of: rm -rf badname, it did rm -rf badname/, which deletes the target of the symlink, which was /.
Obviously this was my fault for using -r for a symlink, since I only wanted to delete one thing
When the command took too long, I got worried, and when I saw ‘can’t delete /sbin/init’, I panicked and aborted it with control+c
Luckily, I had twice daily backups with bacula, to another server. 30 minutes later, everything was restored, and the server didn’t even require a reboot. The 100+ customers on the machine never noticed, since I stopped the rm before it hit /usr/home
There are plenty of other examples of this same problem though
Steam accidently deletes ALL of your files
Bryan Cantrill tells a similiar story from the old SunOS days
Discussion continues and talks about why rm -rf / is blocked by on SunOS and FreeBSD
Additional Coverage: ServerFault
When told to dd the drive to a file, to use testdisk to try to recover files, the user reports accidentally swapping if= and of=, which likely would just error out if the input file didn’t exist, but it might also mean that this entire thing is just a troll. Further evidence: rm -rf / usually doesn’t work on modern linux, without the –no-preserve-root flag