Re: how to wait for nested container launch in tests

2017-07-01 Thread James Peach

> On Jun 30, 2017, at 11:23 AM, Jie Yu  wrote:
> 
> One way is to define a health check?

Good idea, I'll give that a crack

> 
> - Jie
> 
> On Fri, Jun 30, 2017 at 3:36 AM, James Peach  wrote:
> 
>> Hi all,
>> 
>> I'm trying to write a nested container test for an isolator. It turns out
>> that the default executor sends TASK_RUNNING immediately after receiving
>> the LAUNCH_NESTED_CONTAINER response. However, at this time the nested
>> container is not actually up, and the command specified in the task info
>> hasn't been launched. Can anyone suggest a way for a test to detect that
>> the nested container command is actually running, so that I can advance the
>> clock to reliably trigger the isolator?
>> 
>> J



Re: how to wait for nested container launch in tests

2017-06-30 Thread Jie Yu
One way is to define a health check?

- Jie

On Fri, Jun 30, 2017 at 3:36 AM, James Peach  wrote:

> Hi all,
>
> I'm trying to write a nested container test for an isolator. It turns out
> that the default executor sends TASK_RUNNING immediately after receiving
> the LAUNCH_NESTED_CONTAINER response. However, at this time the nested
> container is not actually up, and the command specified in the task info
> hasn't been launched. Can anyone suggest a way for a test to detect that
> the nested container command is actually running, so that I can advance the
> clock to reliably trigger the isolator?
>
> J


how to wait for nested container launch in tests

2017-06-30 Thread James Peach
Hi all,

I'm trying to write a nested container test for an isolator. It turns out that 
the default executor sends TASK_RUNNING immediately after receiving the 
LAUNCH_NESTED_CONTAINER response. However, at this time the nested container is 
not actually up, and the command specified in the task info hasn't been 
launched. Can anyone suggest a way for a test to detect that the nested 
container command is actually running, so that I can advance the clock to 
reliably trigger the isolator?

J