> On June 21, 2019, 1:32 a.m., Till Toenshoff wrote:
> > 3rdparty/libprocess/src/openssl.cpp
> > Lines 808 (patched)
> > <https://reviews.apache.org/r/70749/diff/4/?file=2151450#file2151450line808>
> >
> >     I know you just moved it, but where do these 100ms come from and how 
> > could we be more explicit about that choice?
> >     
> >     I would suggest to use a const with some explaining comment how that 
> > value was chosen - can we please? :D

I'm afraid I have no idea where the 100ms come from. The suggestion sounds 
good, but I think the changes will fit better in a separate review, since 
they're not really related to hostname validation.


> On June 21, 2019, 1:32 a.m., Till Toenshoff wrote:
> > 3rdparty/libprocess/src/openssl.cpp
> > Lines 984-985 (patched)
> > <https://reviews.apache.org/r/70749/diff/4/?file=2151450#file2151450line984>
> >
> >     This sounds like a great idea worth spending 1 more cycle on -- can we 
> > create and reference a ticket that explains this jazz as nicely as we do 
> > here in the code?
> >     
> >     My thought is that being open about this idea in JIRA,  we would 
> > provide more chances of getting community support for it.

Opened https://issues.apache.org/jira/browse/MESOS-9855


> On June 21, 2019, 1:32 a.m., Till Toenshoff wrote:
> > 3rdparty/libprocess/src/openssl.cpp
> > Lines 989 (patched)
> > <https://reviews.apache.org/r/70749/diff/4/?file=2151450#file2151450line989>
> >
> >     Shall we explain why this is the `right` way - aka best practices?

Huh, I actually removed this now: I originally had a look at the OpenSSL 
example code and at RFC6125, but missed that partial wildcards are only 
disallowed in *internationalized* domain names. (and for these, openssl already 
does the correct thing regardless of this flag.)

With this removed, we're a bit more loose than what a web browser would accept, 
but Mesos is not a web browser so that seems fine.


- Benno


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/70749/#review216017
-----------------------------------------------------------


On June 20, 2019, 5:48 p.m., Benno Evers wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/70749/
> -----------------------------------------------------------
> 
> (Updated June 20, 2019, 5:48 p.m.)
> 
> 
> Review request for mesos, Alexander Rukletsov, Joseph Wu, and Till Toenshoff.
> 
> 
> Bugs: MESOS-9809
>     https://issues.apache.org/jira/browse/MESOS-9809
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> This commit introduces a new libprocess SSL flag
> `hostname_validation_scheme`, which can be used to select
> between the previous hostname validation behaviour and a new
> option to use standardized OpenSSL algorithms to handle
> hostname validation as part of the handshake.
> 
> As a nice side-effect, the new scheme gets rid of reverse DNS
> lookups during TLS connection establishment, which used to be
> a common source of hard-to-debug unresponsiveness in Mesos
> components.
> 
> See `docs/ssl.md` in the follow-up commit for details of and
> differences between the schemes.
> 
> 
> Diffs
> -----
> 
>   3rdparty/libprocess/include/process/ssl/flags.hpp 
> f3483f97f93bb29117b2c78f0f2ed9735d9c4b3a 
>   3rdparty/libprocess/src/openssl.hpp 
> 17bec246e516261f8d772f1647c17f092fae82d1 
>   3rdparty/libprocess/src/openssl.cpp 
> e7dbd67913fa8e7fbbf60dee428e7e38895f86ce 
>   3rdparty/libprocess/src/posix/libevent/libevent_ssl_socket.hpp 
> 6ef5a86566af3439cfe0b06ab3576076623f7be0 
>   3rdparty/libprocess/src/posix/libevent/libevent_ssl_socket.cpp 
> 29a1bf71c1df9d80370455a6269ecea0ec4193b0 
> 
> 
> Diff: https://reviews.apache.org/r/70749/diff/4/
> 
> 
> Testing
> -------
> 
> Todo!
> 
> 
> Thanks,
> 
> Benno Evers
> 
>

Reply via email to