You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@singa.apache.org by "Sheng Wang (JIRA)" <ji...@apache.org> on 2015/08/12 05:25:45 UTC

[jira] [Resolved] (SINGA-30) Enhance easy-to-use feature and support concurrent jobs

     [ https://issues.apache.org/jira/browse/SINGA-30?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Sheng Wang resolved SINGA-30.
-----------------------------
    Resolution: Fixed

Have done all these in following tickets:
SINGA-33 Automatically launch a number of processes in the cluster
SINGA-34 Support external zookeeper service
SINGA-38 Support concurrent jobs

> Enhance easy-to-use feature and support concurrent jobs
> -------------------------------------------------------
>
>                 Key: SINGA-30
>                 URL: https://issues.apache.org/jira/browse/SINGA-30
>             Project: Singa
>          Issue Type: New Feature
>            Reporter: Sheng Wang
>
> To let users run singa jobs easily and concurrently, we need to enhance the functionalities zookeeper services and bash scripts. There are several milestones:
> 1. Automatically launch a number of processes in the cluster
>   User only need to provide a global host list that singa can use, instead of providing a host list for each job.
> 2.  Support running multiple singa jobs at the same time
>   We need to ensure that all runtime information is stored in separated zookeeper paths. The first step is to implement a tool that generates a unique job id for each job, and let zookeeper put contents into different paths.
> 3. Provide an easy-to-use script for users to run and kill singa jobs, such as:
> $./singa.sh start -conf=CONF_DIR
> $./singa.sh kill -job=JOB_ID
> We will get these done step by step.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)