amuck-landowner

Need to patch for Heartbleed? Here are a couple quick how-to posts...

Mun

Never Forget
Good job on making a tutorial that doesn't actually do enough. You still need to reissue your SSL as well, which your guides does not show at all.

Mun
 

VPSCorey

New Member
Verified Provider
Just reboot the server to be safe.

I have also heard that you may need to rekey your SSL certificates.
 

DomainBop

Dormant VPSB Pathogen
Need to patch for Heartbleed?

People needed to patch the second  the updates were released but of course not everyone did... some web hosts were slow to update and got their panties in a wad when they were called out for it >> http://www.webhostingtalk.com/showthread.php?t=1364636

yum update -y openssl


service httpd restart.

I'm sure there are many people who just applied the upgrade and didn't restart any services or who thought httpd was the only service that used openssl (case in point, in the linked WHT thread, a host is babbling that his customers weren't at risk from the exploit because his company processes payments offsite at a 3rd party processor...reminds me of the people who think they don't need an SSL certificate because they use PayPal).

So, what is your point for posting exactly? Why didn't you just ask directly the provider if they do update their software or not? Or if they are really using the openssl for their site transactions. What makes you think that our site is doing the payment transactions directly? What makes you think that our customers are on risk?
 

Enterprisevpssolutions

Article Submitter
Verified Provider
You also should check to make sure you don't have any other services running with the old openssl you can do this by running this command

lsof -n | grep ssl | grep DEL

if it returns with nothing your good to go if you have something still using this restarting the service or server will resolve that issue.
 
Last edited by a moderator:

Echelon

New Member
Verified Provider
Easy steps to cover yourself, since there's a lot of misinformation and FUD getting tossed around

1) Update your OpenSSL installations and libraries. If you are unable to move versions, or have a specific way you've compiled OpenSSL for any which way, you can cover your bases by recompiling any affected version with '-DOPENSSL_NO_HEARTBEATS' to disable the vulnerable vector.

2) Revoke and renew your SSL certificates. Regenerate your SSH keys. Change your passwords, and any other credentials. The 'heartbleed' exploit dumps out the contents of the memory in 64K chunks, but this doesn't stop somebody from firing 100,000+ requests down your vulnerable web server to pull down the private keys, passwords, and other data in plain-text sitting pretty. (Ref: https://www.cloudflarechallenge.com/heartbleed)

3) If your affected server houses any systems, control panels, other other mechanisms that customers and other users will be making use of, notify them, and force a password change. Worst case scenario, notify them, and ask them to change their password.

At the end of the day, there hasn't been any public releases of any intrusions resulting from the exploit, but that doesn't mean there won't be, especially considering that it's out in the wild at this point. There's a Windows Sysadmin somewhere in the world that's been getting quite a few good nights' sleep the last few days ;).
 

Wintereise

New Member
You still need to reissue your SSL

and test it here: http://filippo.io/Heartbleed/
Not really. You CAN reissue it if you want to be 100% secure, but many researchers have shown that it's not really very likely that you'll magically stumble upon the entire private key in the heap.

It can happen, but the chances are so slim that you might as well care more about other stuff -- such as Cookie/session data leaks, which actually happens pretty much with every other request.

nginx users are also fairly immune (VERY low address spaces used to store keys), session data is kept much higher -- but eh, to each their own.
 
Top
amuck-landowner