Last Updated: 2014-10-02 14:29:31 UTC
by Johannes Ullrich (Version: 1)
Kuddos to Matthew for paying attention to egress traffic. We keep emphasizing how important it is to make sure no systems talk "outbound" without permission. Just this last week, various Shellshock exploits did just that: Turn devices into IRC clients or downloading additional tools via HTTP, or just reporting success via a simple ping.
So no surprise that Matthew wrote us: "... the first time I saw the storage array SSH to the internet I about fainted. ..."
I would be surprised too! And turns out that isn't the only person that experienced this. Mark noted:
"Had the seem freak moment when I saw it happen. The SAN happily communicating to an outside entity. Though the company had been well and truly hosed."
Luckily, before going too far down the incident handling road, Matthew realized that this was a false positive. The storage array in question called "back home" to the vendor to report on its status. The purpose of this communication is to report failed disks or other critical events that may trigger a service call. Vendors will agree to turn off this feature, but then of course it is up to you to recognize faulty disks.
Got anything like that? Let us know. (if possible with log snippet / packet capture or other show-and-tells)