[jira] [Comment Edited] (BIGTOP-1493) Add smoke test for tachyon

["jay vyas (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/BIGTOP-1493?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14233157#comment-14233157 ] 

jay vyas edited comment on BIGTOP-1493 at 12/3/14 4:21 PM:
-----------------------------------------------------------

- the first tests passes as long as tachyon is set up , which is great.
- for 2nd (mapred) to pass, we will need to implement http://tachyon-project.org/Running-Hadoop-MapReduce-on-Tachyon.html in the puppet recipes.
- I will take that on in BIGTOP-1547.  But this is definetely feature complete in my opinion.

overall looks good. thanks david.   
anyone else have any final thoughts on this?

one minor thought : I think just calculate {{pi 2 2}} is really enough.  i dont know why we need to test two mapreduce jobs.  is there a reason why you have two jobs ?


was (Author: jayunit100):
- I think for this to pass, we will need to implement http://tachyon-project.org/Running-Hadoop-MapReduce-on-Tachyon.html in the puppet recipes.
- I will take that on in BIGTOP-1547.  But this is definetely feature complete in my opinion.

overall looks good. thanks david.   
anyone else have any final thoughts on this?

> Add smoke test for tachyon
> --------------------------
>
>                 Key: BIGTOP-1493
>                 URL: https://issues.apache.org/jira/browse/BIGTOP-1493
>             Project: Bigtop
>          Issue Type: Test
>          Components: tests
>    Affects Versions: 0.8.0
>            Reporter: jay vyas
>            Assignee: David Capwell
>         Attachments: BIGTOP-1493.1.patch, BIGTOP-1493.2.patch, BIGTOP-1493.3.patch, iTest-deps.png
>
>
> Tachyon as an in memory cache will be emerging as a common feature in hadoop and spark based clusters.
> Lets create a smoke test under the {{smoke-tests}} which uses the {{tachyon:///}} url.  
> To get started with the smoke tests, you can 
> 1) Run the vagrant recipe under bigtop-deploy/vm/vagrant-puppet .  This will demonstrate the way you spin up a bigtop based hadoop cluster. 
> 2) Now, you can do {{vagrant ssh}} into your cluster, and simply edit the smoke tests or install stuff.  Re running the smoke-tests.sh file should test that your new tests work .  At that point, just exit your VM, and commit your changes and submit the new patch. 



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