Re: Question about behavior after container failures

2014-11-14 Thread Sumit Mohanty
*If I kill the child process *
So are you observing that a replacement container instance is never
started? Can you check the AM log? What is expected to happen here is as
follows:
- process goes down, and then agent goes down, AM notices, asks for a new
container from Yarn.

if you want the agent to restart the process then you need to set "
 true" for the component in the
metainfo.xml. This is not part of 0.40 though.

*If I kill the agent process*
This is YARN-1922 (https://issues.apache.org/jira/browse/YARN-1922).

On Fri, Nov 14, 2014 at 4:46 PM, hsy...@gmail.com  wrote:

> Hi all,
>
> In the document, I got that slider will try to recover on container
> failure. But in my test application it doesn't
>
> I'm using the 0.40 release. build from source
>
> Here is what I get.
>
> *If I kill the child process *
> The agent check_process_status method will raise ComponentIsNotRunning and
> never get back again
>
> 2014-11-14 16:18:40,274 - Error while executing command 'status':
> Traceback (most recent call last):
>   File
> "/yarn/nm/usercache/vagrant/appcache/application_1415305968048_0008/container_1415305968048_0008_01_02/infra/agent/slider-agent/resource_management/libraries/script/script.py",
> line 114, in execute
> method(env)
>   File
> "/yarn/nm/usercache/vagrant/appcache/application_1415305968048_0008/container_1415305968048_0008_01_02/app/definition/package/scripts/kafka.py",
> line 60, in status
> check_process_status(status_params.pid_file)
>   File
> "/yarn/nm/usercache/vagrant/appcache/application_1415305968048_0008/container_1415305968048_0008_01_02/infra/agent/slider-agent/resource_management/libraries/functions/check_process_status.py",
> line 45, in check_process_status
> raise ComponentIsNotRunning()
> ComponentIsNotRunning
>
>
> *If I kill the agent process*, it's just silently gone
>
>
>
> Best,
>
> Siyuan
>

-- 
CONFIDENTIALITY NOTICE
NOTICE: This message is intended for the use of the individual or entity to 
which it is addressed and may contain information that is confidential, 
privileged and exempt from disclosure under applicable law. If the reader 
of this message is not the intended recipient, you are hereby notified that 
any printing, copying, dissemination, distribution, disclosure or 
forwarding of this communication is strictly prohibited. If you have 
received this communication in error, please contact the sender immediately 
and delete it from your system. Thank You.


Question about behavior after container failures

2014-11-14 Thread hsy...@gmail.com
Hi all,

In the document, I got that slider will try to recover on container
failure. But in my test application it doesn't

I'm using the 0.40 release. build from source

Here is what I get.

*If I kill the child process *
The agent check_process_status method will raise ComponentIsNotRunning and
never get back again

2014-11-14 16:18:40,274 - Error while executing command 'status':
Traceback (most recent call last):
  File 
"/yarn/nm/usercache/vagrant/appcache/application_1415305968048_0008/container_1415305968048_0008_01_02/infra/agent/slider-agent/resource_management/libraries/script/script.py",
line 114, in execute
method(env)
  File 
"/yarn/nm/usercache/vagrant/appcache/application_1415305968048_0008/container_1415305968048_0008_01_02/app/definition/package/scripts/kafka.py",
line 60, in status
check_process_status(status_params.pid_file)
  File 
"/yarn/nm/usercache/vagrant/appcache/application_1415305968048_0008/container_1415305968048_0008_01_02/infra/agent/slider-agent/resource_management/libraries/functions/check_process_status.py",
line 45, in check_process_status
raise ComponentIsNotRunning()
ComponentIsNotRunning


*If I kill the agent process*, it's just silently gone



Best,

Siyuan