Re:Synology Security Advisory
by landgenoot on Thursday November 19, 2026 @08:00PM
(#50568825)
Applications are vulnerable for exploits.
mTLS prevents connecting to the application without valid cert.
Sure the reverse proxy handling the mtls can be exploited, but so can any VPN server as well.
authn and authz should be separate. mTLS is used first for identity attestation, what you’re allowed to do after you’ve proven the identity is authz.
Re:Synology Security Advisory
by catlifeonmars on Friday November 20, 2026 @12:21AM
(#50570483)
[delayed]
Yeah that I get, but I’m disputing that the tool (mTLS) is not at fault for being misused, authz (rather the lack of it) is being misused. The tool (authn) is working fine.
Re:Synology Security Advisory
by cloudfudge on Friday November 20, 2026 @03:08AM
(#50571592)
We are saying the same thing. mtls is fine for authn. But nobody actually uses it for just that. They use it for both authn and authz. The authz policy is almost always just "identity != null".
If I told you not to use a power drill as a hammer, you wouldn't claim I'm saying the hammer is at fault for being a bad power drill.
Re:Synology Security Advisory
by catlifeonmars on Friday November 20, 2026 @03:40AM
(#50571861)
[delayed]
Yes, this I agree with that. JWT is simpler to work with by virtue of remaining on http layer 7 and parsing claims out of JSON compared to working with extracting identity and other embeddings from x509.
I just take issue (maybe pedantically so) calling it a bad tool and blanket statements that no one uses it for just authn but for authz also. It’s not used for authz. The absence of authz shouldn’t be blamed on mTLS.
Now if you're going to use mTLS as the basis for mature authn/authz, that's a different story. But I never see anyone do that.