[
https://issues.apache.org/jira/browse/HADOOP-11892?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15238299#comment-15238299
]
Wei-Chiu Chuang commented on HADOOP-11892:
------------------------------------------
[~xiaochen] You got it right. And you can also use lsof /dev/urandom to list
open fds.
However, I have a different opinion on the fix. When we typically see problems
with too many open fds to /dev/urandom, it's not the cost of instantiating a
new instance of {{CryptoCodec}} that is expensive; it is generating random
numbers that is expensive. The problem is when available entropy is depleted,
reading from /dev/urandom could block, and what happens is you would end up
seeing many open fds, because they are all waiting to read from /dev/urandom.
IMO, fixing this isn't really useful.
> CryptoCodec#getInstance always returns a new instance of CryptoCodec. This
> could be expensive
> ---------------------------------------------------------------------------------------------
>
> Key: HADOOP-11892
> URL: https://issues.apache.org/jira/browse/HADOOP-11892
> Project: Hadoop Common
> Issue Type: Bug
> Affects Versions: 2.7.0
> Reporter: Arun Suresh
> Assignee: Arun Suresh
>
> {{CryptoCodec#getInsatnce}} should be able to return possibly cached
> instances of the CryptoCodec implementation as instantiating a new instance
> every time could be expensive
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
