[ 
https://issues.apache.org/jira/browse/HIVE-15688?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15852497#comment-15852497
 ] 

Siddharth Seth commented on HIVE-15688:
---------------------------------------

If I'm reading this correctly, when the "startImmediately" option is specified, 
the following happens
1) Normal processing in the JavaServiceDriver
2) This then invokes package.py with a specific OUTPU_DIR, and invokes the 
start.
3) The java processing falls off, and we run package.py again - which generates 
the same contents as what we do today.

Is that correct?

On the patch itself
{code}
scriptArgs.add(OUTPUT_DIR_PREFIX + version);
{code}
Should this be to a temporary directory?

Move startCluster into SliderUtils?

{code}
if (rc == 0) {
          startCluster(options.getName(), version,
              HiveConf.getVar(conf, ConfVars.LLAP_DAEMON_QUEUE_NAME));
        }
{code}
Some error checking  / message returned from calling python would be useful in 
case this fails.


Even if we run via this, was hoping we'd generate a run.sh which could invoke 
java instead of the command line slider tools. Re-strating with the same 
settings happens quite often while testing.

Unrelated to this jira: Any idea why package.py is in python. Is this something 
that can be moved to Java later, so that we don't have logic split in 2 places?

> LlapServiceDriver - an option to start the cluster immediately
> --------------------------------------------------------------
>
>                 Key: HIVE-15688
>                 URL: https://issues.apache.org/jira/browse/HIVE-15688
>             Project: Hive
>          Issue Type: Bug
>            Reporter: Sergey Shelukhin
>            Assignee: Sergey Shelukhin
>         Attachments: HIVE-15688.01.patch, HIVE-15688.patch
>
>
> run.sh is very slow because it's 4 calls to slider, which means 4 JVMs, 4 
> connections to RM and other crap, for   2-5sec. of overhead per call, 
> depending on the machine/cluster.
> What we need is a mode for llapservicedriver that would not generate run.sh, 
> but would rather run the cluster immediately by calling the corresponding 4 
> slider APIs. Should probably be the default, too. For compat with scripts we 
> might generate blank run.sh for now.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Reply via email to