> On April 11, 2017, 12:27 a.m., Vinod Kone wrote:
> > src/authorizer/local/authorizer.cpp
> > Lines 725 (patched)
> > <https://reviews.apache.org/r/58254/diff/2/?file=1686563#file1686563line725>
> >
> >     Is this based on the assumption that claims subjects only come from 
> > executors and not operators? What guarantees that?

There's one use case this patch would not accommodate: if a custom 
authenticator is used which sets both `Principal.value` and `Principal.claims`, 
and the local authorizer is also used. In that case, an operator could 
authenticate such that this code would not authorize their request correctly. 
To address this, I could add a check here for `!subject->has_value()`, since 
only implicit executor authZ can handle subjects without a value, and the 
default JWT authenticator does not set `Principal.value`.


- Greg


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/58254/#review171506
-----------------------------------------------------------


On April 7, 2017, 11:25 p.m., Greg Mann wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/58254/
> -----------------------------------------------------------
> 
> (Updated April 7, 2017, 11:25 p.m.)
> 
> 
> Review request for mesos, Adam B, Alexander Rojas, Till Toenshoff, and Vinod 
> Kone.
> 
> 
> Bugs: MESOS-7014
>     https://issues.apache.org/jira/browse/MESOS-7014
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> This patch updates the agent handlers for the LAUNCH_, WAIT_,
> and KILL_NESTED_CONTAINER calls of the operator API to set the
> `container_id` field within the authorization object,
> facilitating implicit executor authorization.
> 
> 
> Diffs
> -----
> 
>   include/mesos/authorizer/authorizer.proto 
> 736f76d552956f2351ffd40fc51d088dff83f8c8 
>   src/authorizer/local/authorizer.cpp 
> e241edf4afa48d35dbbbb94d72e8e8690f5bedfc 
>   src/slave/http.cpp b07ce7c73a90ef297d980806ebba9530d86f25ae 
> 
> 
> Diff: https://reviews.apache.org/r/58254/diff/2/
> 
> 
> Testing
> -------
> 
> Testing details can be found at the end of this chain.
> 
> 
> Thanks,
> 
> Greg Mann
> 
>

Reply via email to