Hansen: SKS Keyserver Network Under Attack

What prevents uploading the same malicious certificates to other keyservers? Is there any reason to believe (or know) that other keyserver networks haven't experienced similar compromise?

"At present I (speaking only for myself) do not believe the global keyserver network is salvageable. High-risk users should stop using the keyserver network immediately."

And I'm truly sorry to read of the emotional effect that this thoughtless act has had on one of the early and prolific contributors of the project. My hope is that one or two or ten of the large multinational corporations that now feed off of volunteer labor of F/OSS projects will step up to put real resources behind fixing the design, usability and implementation problems detailed in the article.

Eventually we need to remove the search by user id feature which is anyway a bad idea because too many users simply trust what they get back from the keyservers. Further, the Web-of-Trust is mostly a hacker's game and I guess we can do fine without and thus we can also drop any public key-signatures and thus remove a major DoS vector.

Keyservers are still useful to distribute revocations and I hope that the volunteers running the network will eventually install new software which allows just the distribution and easy access of revocations.

"The software is unmaintained. Due to the above, there is literally no one in the keyserver community who feels qualified to do a serious overhaul on the codebase."

"The keyserver network is susceptible to a variety of attacks as a consequence of its write-only design... We have known about these vulnerabilities for well over a decade. Fixing the keyserver network is, however, problematic for the reasons listed above."

As far as I'm concerned, everyone involved that knows about this is doing everyone who uses such keyservers a major disservice by not making that public knowledge, loud and clear.

Not to mention:

"These attestations — what we call certificate signatures — can be made by anyone for any purpose. And once made, they never go away. Ever. The OpenPGP specification puts no limitation on how many signatures can be attached to a certificate. The keyserver network handles certificates with up to about 150,000 signatures. GnuPG, on the other hand … doesn't. Any time GnuPG has to deal with such a spammed certificate, GnuPG grinds to a halt. It doesn't stop, per se, but it gets wedged for so long it is for all intents and purposes completely unusable."

Which has nothing to do with the keyservers or their archaic unmaintainable software, but GnuPG itself, and the design of the system.

"We've known for a decade this attack is possible. It's now here and it's devastating." "At present I (speaking only for myself) do not believe the global keyserver network is salvageable. High-risk users should stop using the keyserver network immediately."

Then we should already be using keyserver-protocol-v2 (whatever that may be, and however incompatible it has to be made), and should have started work on it ten years ago.

There's an attack on the person who perpetrated this attack, calling them a fool at the bottom. I feel there should be an entirely different target here - all the people involved in such keyserver operaton and utilisation who were aware of this kind of attack who predicated such simple things as "everyday software updates" on them without ever bothering to replace them with something better the second this was known.

Turn the damn keyservers off now, deal with the carnage created, because having a poison certificate that belongs in a major mainstream package will bring *EVERYONE'S* update process to a grinding halt and put us all at serious risk... combine that attack on the right package/person with a major vulnerability in, say, Apache or SSH and you can take down almost everyone without giving them a way to update in any sensible time.

This is far more serious than an article on LWN.net. This is "switch it off!". The only time I've ever had to yell that is when there's been an actual fire, or when I suspect malware on a network (the response to which is "switch off everything suspected to be affected immediately, survey the impact, isolate, and check before bringing it back online")

Another comment mentions WKS. I knew nothing about this until I followed the link and did a bit more searching. I came upon this blog entry from Werner in 2016: https://www.gnupg.org/blog/20160830-web-key-service.html

Already I think this is not an option for me as the managed hosting I use does not allow me root access, so the steps he provides are moot to me.

It appears then that WKS/WKD is not an exact replacement for the SKS server network? Without using the public key servers, what are we end users to do? Do I disable keyserver lookups entirely? Do I even understand a way to mitigate the potential problem?