Cauterizing OpenSSL's heartbleed (the aftermath)

["mancha" <mancha1 () hush com>]

There's an important distinction between systems "never vulnerable to
heartbleed" (for appropriate definitions of never) and those fixed in
response to security announcements. The latter category should
definitely be prioritizing revocation/reissuance of certificates and
other potentially compromised credentials.

Mustafa Al-Bassam's work assists a great deal with this taxonomy. He
ran PoC code against Alexa top 100, 1000, and 10000 sites beginning
about 18 hours after OpenSSL's first public announcement [1].

Specifically, his scans began circa: 1396956600 (top 100); 1396958400
(top 1000); and 1396972800 (top 10000). Did any major vendors deploy
upgrades prior to this?

If others have done similar work (hopefully closer to time zero), please
share to complement Mustafa's good work.

--mancha

====

[1] https://github.com/musalbas/heartbleed-masstest

-----------------
PGP: 0x25168EB24F0B22AC
[56B7 100E F4D5 811C 8FEF  ADD1 2516 8EB2 4F0B 22AC]