hpr-knowledge-base/hpr_transcripts/hpr1491.txt

Episode: 1491
Title: HPR1491: Heartbleed
Source: https://hub.hackerpublicradio.org/ccdn.php?filename=/eps/hpr1491/hpr1491.mp3
Transcribed: 2025-10-18 04:08:22

---

Here goes!
This is Landeer for Hacker Public Radio, recording Friday, April 11th, 2014, Heartblade.
Once again, I am recording from my car on the drive home from work, so I apologize for the
recording quality.
I listened to my last one and I realized there's a lot of environmental sound, but please
bear with me.
Today I'd like to talk to you about the OpenSSL Heartblade Vulnerability, officially CVE-2014-0160.
I recognize that many of you have probably already heard of the vulnerability.
It has gotten a lot of media coverage, particularly among the types of tech blogs and podcasts,
which have a very large overlap with the Hacker Public Radio audience.
But for those of you who have not heard or have not decided on any action to take, I'd
just like to briefly go over what the vulnerability is, what kinds of exposure have gone on,
and some possible courses of action you can take to help protect yourself.
So firstly, what is Heartblade?
Heartblade is essentially a bounce-checking error in Lib OpenSSL in OpenSSL's TLS implementation.
And code for what's known as SSL Heartbeats.
Now Heartbeats are an extension to the SSL SPAC, basically designed as a keep-alive.
Now I'm not sure why these would be widely used in SSL over TCP since TCP has its own keep-alives,
but it's there, it's in use, many servers support it, so it's something we need to be aware of.
Now, from my understanding, the Heartbeats extension was introduced in OpenSSL 1.0.1,
and accordingly that is the first version, which is vulnerable to the Heartblade attack.
Now, I guess before I go any further, I should mention that the patch was introduced in 1.0.1G,
and so the vulnerable versions are 1.0.1 through 1.0.1F and 1.0.2 beta.
Now, if you're on Debian Stable, like I am, where packages tend to be feature-frozen,
you may have a distribution-specific patch level applied to backport the security fix.
Debian Stable, for example, as of this recording, is at 1.0.1E-2DebShortForDebian,
and some other numbers that I cannot recall right at this moment.
But if you look at the changelog.debian.gz in the documentation folder from OpenSSL,
you will see CVE 2014 0160 listed as fixed.
Now, I've heard some varying reports in how widespread Heartblade vulnerability has been,
but the most recent estimate I heard puts it at somewhere around 60% of internet servers,
maybe higher, maybe lower, but it's extremely widespread.
Additionally, because of when this bug was introduced, it's been in existence
and theoretically exploitable for over two years.
So, really, this is quite a serious thing.
How serious exactly we don't really know.
So, back to exactly how the bug works.
In the Heartbeat extension code, there is a function which is responsible for generating
the Heartbeat response back to the client.
And really, what happens is that it reads a length right off the wire coming from the client,
and eventually down the line passes that into a call to the Unix Memcopy function.
What that means is that the data that goes into the response reads beyond what the authors of the code intended
and goes off arbitrarily somewhere into the process is memory.
Now, based on how open SSL is actually being used and what kind of allocations are being done,
it's really anybody's guess as to what that memory might contain.
The important thing to note is that it could contain anything that the open SSL process is already using.
So, let's take the example of a web server such as Apache running, say, some,
I don't know, some kind of a website that takes a login.
Since we're open over open SSL, Apache has to have the open SSL code loaded.
It has to hold at any given time in memory the open SSL private key in order to authenticate itself to new clients.
It has to hold any data used internally by Apache that could include information from the file system,
it could include information from applications running as modules inside Apache,
and probably most shocking of all is that it could contain data that's being read off the wire and decrypted.
Parts of HTTP sessions for actual HTTP users, people accessing this website,
which means this could contain anything as benign as, say, the user agent string that identifies what version the browser is
to the URL being requested, what page on the site, but perhaps to give you an idea of the scope of what's going on,
if you enter your password on a site, in most cases, that's going to be sent over the wire.
Now, SSL would normally encrypt it, but once it gets to the web server, it has to be decrypted at the other end.
So the application can do whatever it needs to do with that password.
So memory could contain your clear text passwords, it could contain session cookies that could be used to hijack a session.
So if you can remember back to the fire sheep extension, which was used to grab things like social media, login cookies,
from other people's sessions on unprotected Wi-Fi, which would allow them to impersonate those users.
So even if they don't get your password, they still could potentially log in as you.
Also, think about banking sites.
This could potentially have other personal data, bank account numbers, balances, in a worst case, social security numbers,
or any other kind of information that you really wouldn't want to get out.
So that's the kinds of things that can get away.
Now, let's say that you use a website, and their administrator is really on top of things.
They update to the latest version of open SSL to close this heartbeat bug.
That's all well and good.
Your data can no longer get out through that hole, it's been plugged.
Here's the problem.
Let's suppose that during the window that that website was vulnerable to heart bleed, someone managed to dump some memory,
which contained the server's certificate private key.
That key can be used to impersonate the server.
So if you maybe were the target of a DNS poisoning or some kind of other men in the middle attack,
someone could set up a server between you and your banking site.
They would be able to successfully authenticate as your banking site.
It would look like they validly have the correct certificate because they do have the private key for most recently received certificate.
And they can then sit between you and your bank, look at anything you send through decrypting traffic on the fly,
and showing you anything they want to.
And it'll look to you exactly like it's coming from your bank.
So if that weren't bad enough, depending on the kind of SSL configuration your site is using,
let's say someone has a packet capture of a previous session where they weren't set up as a man in the middle to actually decrypt the traffic,
but they were just able to capture it. So maybe this was on an open Wi-Fi,
but you were still accessing it over what you thought was a secure connection.
So they have this packet capture, and then at some point later they get the private keys.
Depending on their SSL configuration at that site, they might be able to use that key to retroactively decrypt that entire captured session.
That is one of the reasons it is important to configure SSL to use what are known as ephemeral keys.
You may also hear this referred to as perfect forward secrecy, meaning that once the session is done,
there's nothing kept around that can be used to decrypt that session later.
So bare minimum, what a site needs to do if they were vulnerable to this bug.
They need to switch to a version of open SSL that is not vulnerable, so preferably 1.0.1 G or later,
or the appropriate distra-specific version. They need to obtain a new certificate,
as any certificates they had private keys for on that server could have been compromised.
They should issue a revocation notice for the old certificate,
and once those things are done, traffic to and from that website should again be secure.
Hopefully they'll also take that opportunity to enable perfect forward secrecy or ephemeral keys.
Now, if they've done all that, it's still possible that while you were using the site during a vulnerable period,
some of your data could have been leaked through heartbleed or some data that you sent or received inside a session that used a compromised certificate could have been intercepted.
So any information on that site that belongs to you is theoretically vulnerable.
If you have banking or financial information, it may be a good idea to just keep an eye on your accounts for a while,
just watch for any suspicious activity.
There have not yet been any widespread reports of ongoing heartbleed attacks in the wild,
but it's only a matter of time, at least until these sites can get updated.
So, would strongly recommend keeping an eye on sensitive information that could have impacts outside the site.
Within the site, go ahead and change your password, update any other credentials that could be used to gain access to your account.
If you use the same password on multiple sites, obviously this is even more important.
Maybe take this time to update passwords on all of your sites, consider using a password manager such as last pass or not as familiar with the open source ones.
I believe the name I'm searching for is key pass or key pass x.
And just use that as a tool that will help you to securely maintain unique passwords across multiple sites.
So, that's something you can do just as an end user to keep yourself safe.
Again, hold off on making those changes until you're sure that the server vulnerabilities have been patched,
that they're using a new version of open SSL and that their certificates have been updated.
Most of the websites that I use on a regular basis have been extremely forthcoming with how they were affected and what steps they've taken to correct it.
So, hopefully with that in mind, you'll be able to keep yourself safe in light of the heartbeat vulnerability.
One other note, just for those who may have come across some confusion, this is a bug in open SSL.
This is not at all affecting open SSH.
There are two totally different protocols.
The heartbeat feature is completely open SSL or I'm sorry SSL specific, so you do not need to be concerned about that with open SSH.
So, again, this comes back to website usage over SSL or TLS.
There are other protocols that do use SSL TLS, SMTP for email, some VPN technologies, but obviously the most prevalent is HTTPS.
You have been listening to Hacker Public Radio at Hacker Public Radio.
We are a community podcast network that releases shows every weekday Monday through Friday.
Today's show, like all our shows, was contributed by a HPR listener like yourself.
If you ever consider recording a podcast, then visit our website to find out how easy it really is.
Hacker Public Radio was founded by the digital dog pound and the infonomicum computer club.
HPR is funded by the binary revolution at binref.com, all binref projects are proudly sponsored by Lina pages.
From shared hosting to custom private clouds, go to LinaPages.com for all your hosting needs.
Unless otherwise stasis, today's show is released under a creative commons, attribution, share a life, lead us our license.
Hacker Public Radio was founded by the digital dog pound and the infonomicum computer club.
Hacker Public Radio was founded by the digital dog pound and the infonomicum computer club.
Hacker Public Radio was founded by the digital dog pound and the infonomicum computer club.
Hacker Public Radio was founded by the digital dog pound and the infonomicum computer club.