> On Nov. 13, 2015, 9:59 a.m., Joris Van Remoortere wrote:
> > Can we fix the underlying problem, as opposed to disabling SSL?
> 
> Jojy Varghese wrote:
>     Not sure if we should enforce SSL sockets in an executor unless being 
> explicitly asked for (say using a flag)

I don't think hard coding this is the right way to fix this, and this is going 
to cause more problems as it affects all executors.
We should investigate how to get SSL to work when the environment is present, 
and also have a better way to control all components either SSL enabled or not.


- Timothy


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/40284/#review106388
-----------------------------------------------------------


On Nov. 13, 2015, 8:12 a.m., Jojy Varghese wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/40284/
> -----------------------------------------------------------
> 
> (Updated Nov. 13, 2015, 8:12 a.m.)
> 
> 
> Review request for mesos and Timothy Chen.
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> When SSL is enabled, the executor process will also expect SSL certs, keys 
> etc.
> which causes unnecessary dependencies. A simple case is when password is
> required for the SSL key. In this case, the executor would launch and wait for
> user on terminal to enter password.
> 
> 
> Diffs
> -----
> 
>   src/slave/containerizer/mesos/containerizer.cpp 
> 08243b61c1c277da7609bc910323cc1e27ff5cd4 
> 
> Diff: https://reviews.apache.org/r/40284/diff/
> 
> 
> Testing
> -------
> 
> 
> Thanks,
> 
> Jojy Varghese
> 
>

Reply via email to