[
https://issues.apache.org/jira/browse/PIG-2672?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13879989#comment-13879989
]
Koji Noguchi commented on PIG-2672:
-----------------------------------
Thanks Aniket! I like the non-share approach.
bq. (to avoid cleanups).
Had a discussion with [~jlowe] about this before. In our cluster, longest job a
user can run on our cluster is 1 week. (due to delegation token limit we set).
With this assumption, we can create a cache under .Trash as below.
<noformat>
$ echo /user/$USER/.Trash/$(date -d 'next monday + 1week' +'%y%m%d'000000)
/user/knoguchi/.Trash/140203000000 (this is in 0.23/2.* format. 0.20 uses
slightly different format)
<noformat>
This way, files are reused for 1 week and then thrown away later automatically
by a Trash cleanup.
We threw away the idea for various reasons, but just wanted to share here.
> Optimize the use of DistributedCache
> ------------------------------------
>
> Key: PIG-2672
> URL: https://issues.apache.org/jira/browse/PIG-2672
> Project: Pig
> Issue Type: Improvement
> Reporter: Rohini Palaniswamy
> Fix For: 0.13.0
>
> Attachments: PIG-2672-5.patch, PIG-2672.patch
>
>
> Pig currently copies jar files to a temporary location in hdfs and then adds
> them to DistributedCache for each job launched. This is inefficient in terms
> of
> * Space - The jars are distributed to task trackers for every job taking
> up lot of local temporary space in tasktrackers.
> * Performance - The jar distribution impacts the job launch time.
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
