> On July 17, 2017, 12:23 p.m., James DeFelice wrote:
> > Some questions about this API:
> > 
> > 1. If a launch request times out, how might someone query the state of the 
> > requested launch operation/container?
> > 2. When a stand alone container terminates (either success or failure), who 
> > is notified and when?
> > 3. What are the expected failure modes/codes for this API and what are the 
> > recovery semantics?

Without this follow-up feature 
(https://issues.apache.org/jira/browse/MESOS-7492 ), standalone containers will 
basically be launch-and-forget.

In fact, I may want to discard this API entirely as the DaemonManager component 
will supposedly have a more comprehensive API for launching standalone 
containers.


- Joseph


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/60890/#review180722
-----------------------------------------------------------


On July 14, 2017, 5:41 p.m., Joseph Wu wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/60890/
> -----------------------------------------------------------
> 
> (Updated July 14, 2017, 5:41 p.m.)
> 
> 
> Review request for mesos, Gilbert Song and Jie Yu.
> 
> 
> Bugs: MESOS-7305
>     https://issues.apache.org/jira/browse/MESOS-7305
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Launching a standalone container is very similar to launching a
> nested container, except that the caller does not specify a
> ContainerID (which will be auto-generated) and must specify some
> Resources.  The CommandInfo and ContainerInfo fields behave similarly.
> 
> 
> Diffs
> -----
> 
>   include/mesos/agent/agent.proto 9bac9541acd24e1123ca5dd5925e2a1381d13b4a 
>   include/mesos/v1/agent/agent.proto ea9282cf12fbe1c2ddeaa37223e4811685263734 
> 
> 
> Diff: https://reviews.apache.org/r/60890/diff/1/
> 
> 
> Testing
> -------
> 
> make
> 
> See later patches in chain.
> 
> 
> Thanks,
> 
> Joseph Wu
> 
>

Reply via email to