https://caddy.community/t/random-ocsp-response-errors-for-random-clients/2473?u=matt
Certificates are keyed by name in the cache, optimized for fast lookups
during TLS handshakes using SNI. A more "correct" way that is truly a
1:1 would be to cache certificates by a hash of the leaf's DER bytes,
but this involves an extra index to maintain. So instead of that, we
simply choose to prevent overlap when keying certificates by server
name. This avoids the ambiguity when updating OCSP staples, for instance.
* introduced own ChallengeProvider type, based on acme.ChallengeProvider to avoid vendoring/version mismatches in Caddy plugins; see Caddy issue #1697
* fixed up comments for ChallengeProvider
* moved ChallengeProvider to caddytls/tls.go
* Disable warning for insecure CA if located on private network.
* Add IsPrivateNetwork function
* Add tests
Signed-off-by: Jonas Östanbäck <jonas.ostanback@gmail.com>
* Add more testcases
Signed-off-by: Jonas Östanbäck <jonas.ostanback@gmail.com>
* Rename IsPrivateNetwork -> IsInternal
Signed-off-by: Jonas Östanbäck <jonas.ostanback@gmail.com>
This could have just as easily been a tls directive property in the
Caddyfile, but I figure if these challenges are being disabled, it's
because of port availability or process privileges, both of which would
affect all sites served by this process. The names of the flag are long
but descriptive.
I've never needed this but I hear of quite a few people who say they
need this ability, so here it is.
This commit removes _almost_ all instances of hard-coded ports 80 and
443 strings, and now allows the user to define what the HTTP and HTTPS
ports are by the -http-port and -https-ports flags.
(One instance of "80" is still hard-coded in tls.go because it cannot
import httpserver to get access to the HTTP port variable. I don't
suspect this will be a problem in practice, but one workaround would be
to define an exported variable in the caddytls package and let the
httpserver package set it as well as its own HTTPPort variable.)
The port numbers required by the ACME challenges HTTP-01 and TLS-SNI-01
are hard-coded into the spec as ports 80 and 443 for good reasons,
but the big question is whether they necessarily need to be the HTTP
and HTTPS ports. Although the answer is probably no, they chose those
ports for convenience and widest compatibility/deployability. So this
commit also assumes that the "HTTP port" is necessarily the same port
on which to serve the HTTP-01 challenge, and the "HTTPS port" is
necessarily the same one on which to serve the TLS-SNI-01 challenge. In
other words, changing the HTTP and HTTPS ports also changes the ports
the challenges will be served on.
If you change the HTTP and HTTPS ports, you are responsible for
configuring your system to forward ports 80 and 443 properly.
Closes#918 and closes#1293. Also related: #468.
* tls: Refactor TLS config innards with a few minor syntax changes
muststaple -> must_staple
"http2 off" -> "alpn" with list of ALPN values
* Fix typo
* Fix QUIC handler
* Inline struct field assignments
The loop which performs renewals in the background obtains a read lock
on the certificate cache map, so that it can be safely iterated. Before
this fix, it would obtain the renewals in the read lock. This has been
fine, except that the TLS-SNI challenge, when invoked after Caddy has
already started, requires adding a certificate to the cache. Doing this
requires an exclusive write lock. But it cannot obtain a write lock
because a read lock is obtained higher in the stack, while the loop
iterates. In other words, it's a deadlock.
I was able to reproduce this issue consistently locally, after jumping
through many hoops to force a renewal in a short time that bypasses
Let's Encrypt's authz caching. I was also able to verify that by queuing
renewals (like we do deletions and OCSP updates), lock contention is
relieved and the deadlock is avoided.
This only affects background renewals where the TLS-SNI(-01) challenge
are used. Users report seeing strange errors in the logs after this
happens ("tls: client offered an unsupported, maximum protocol version
of 301"), but I was not able to reproduce these locally. I was also not
able to reproduce the leak of sockets which are left in CLOSE_WAIT.
I am not sure if those are symptoms of running in production on Linux
and are related to this bug, or not.
Either way, this is an important fix. I do not yet know the ripple
effects this will have on other symptoms we've been chasing. But it
definitely resolves a deadlock during renewals.
By calling SetTLSAddress, the acme package reset the challenge provider
to the default one instead of keeping the custom one we specified before
with SetChallengeProvider. Yikes. This means that Caddy would try to
open a listener on port 443 even though we should have been handling it
with our provider, causing the challenge to fail, since usually port 443
is in use.
So this change just reorders the calls so that our provider takes
precedence.
cf. https://github.com/xenolf/lego/pull/292
We renamed caddytls.ErrStorageNotFound to caddytls.ErrNotExist to more
closely mirror the os package. We changed it to an interface wrapper
so that the custom error message can be preserved. Returning only "data
not found" was useless in debugging because we couldn't know the
concrete value of the error (like what it was trying to load).
Users can do a type assertion to determine if the error value is a "not
found" error instead of doing an equality check.
A Caddyfile using *.example.com as its site address would be subject to
this bug at renewal time, as it would use the literal "*.example.com"
value instead of the name being passed in to obtain a certificate.
This change fixes the LoadSite call so that it looks in the proper
directory for the certificate resources.
It was set by default on the caddy-internal config object, and even
checked for conflicts, but it was never actually reflected on the
tls.Config.
This will have user-visible changes: a client that prefers, say, AES-CBC
but also supports AES-GCM would have used AES-CBC befor this, and will
use AES-GCM after.
This is desirable and important behavior, because if for example the
server wanted to support 3DES, but *only if it was strictly necessary*,
it would have had no way of doing so with PreferServerCipherSuites
false, as the client preference would have won.
