How worried about this should I be? Are there plaintext passwords exposed, or do they just have a lot of properly salted hashes that aren't much use to an attacker?
pbkdf2_sha256$12000$zhMKabMgayvK$iniviUCcX9y2PYJcm0AoB3MhybRA1z2Cec1DZnLWxWc=
I think cracking difficulty depends on how many "iterations" they use though.
hash_func$iterations$salt$derived_key
One person estimated an 8-GPU cracking machine two years ago at about 539 billion hashes per minute. At 128k hashes for one password, you could make about 70,182 attempts per second.
But here[1] is a five-machine cluster from a year and a half ago with 25 GPUs. Its speed? 63 billion per second against SHA1. This results in 492,187 attempts per second. Assuming SHA256 is about 50% slower, this would be around 246,093 per second.
Some password dictionaries contain millions of words. But if your password is '0Password', it'll probably be cracked in a couple of seconds on modern hardware.
[1] http://arstechnica.com/security/2012/12/25-gpu-cluster-crack...
However, if you were targeting a specific user and they didn't use a particularly strong password, it's possible that you could brute force it.
cipher/hash: pbkdf2_sha256
cost factor: 12000
salt: zhMKabMgayvK
hash: iniviUCcX9y2PYJcm0AoB3MhybRA1z2Cec1DZnLWxWc=
This exact technique (pretty much) is described here: http://exyr.org/2011/hashing-passwords/. It's a decent, secure way to hash passwords.
Cost factor of 12000 seems solid to me (depends on the hardware they're running on but I'd say brute forcing your way through that would be pretty impossible)
A PBKDF2 cost factor/iteration count of 12000 and 32-byte output means each candidate passphrase costs 12002 SHA256 blocks.
I can buy a crappy bitcoin miner which will do 2GH/s for about USD19.
Let's say we're going to use the Gawker leak as our dictionary. That's ~200,000 candidate passwords.
For a given user, I can therefore find their password (if it exists in the Gawker set) in 12002 * 200000 = 2.4GH SHA256 applications. That will take 1.2 seconds.
So for all 125 million eBay users, that's about 4 years. This work is trivially parallelisable, so buying more or faster hardware is brutally effective.
Note: there is obviously, and hopefully, a non-negligible probability that a user's password isn't in that set. Brute force of (say) the whole 8 printable-ASCII character password space would take longer but would be guaranteed to find to find about 50% (from Adobe leak) of user's passwords.
<algorithm>$<cost>$<salt>/<encrypted_password>
Rainbow tables can't be used against the passwords, so each one will need to be computed individually to either find the result or a collision. That's likely why the seller is only asking for $1000.
Since these are salted and require 12000 iterations, cracking individual passwords will be quite time consuming. The preferred method in this case, though, is to go after low hanging fruit.
The way one would do this is to try something like the 500 most common passwords against all entries in the table. This won't take very long (compared to trying to brute force a bunch of individual passwords), and will probably yield a ton of passwords.
EDIT: I stand corrected, see the reply by sp332 below.
Plaintext uniquely identifying information like Date of Birth was included, however.
Plus you already have your email publicly displayed here ... and i found some weird stuff about amateur ... xxx movies when i googled you T_T
Also, you're an idiot.