You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@trafodion.apache.org by "David Wayne Birdsall (JIRA)" <ji...@apache.org> on 2015/11/05 19:52:27 UTC

[jira] [Commented] (TRAFODION-1591) Need a better way to pick ports for install_local_hadoop

    [ https://issues.apache.org/jira/browse/TRAFODION-1591?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14992239#comment-14992239 ] 

David Wayne Birdsall commented on TRAFODION-1591:
-------------------------------------------------

Today Trafodion developers in workstation environments often use "install_local_hadoop" to install a working copy of Hadoop and HBase on their workstation. Since workstations are often shared among several developers, one cannot typically use the default port assignments. The install_local_hadoop script supports the specification of alternative port assignments.

Today, install_local_hadoop offers three ways to assign ports: -p fromDisplay will pick a port range that is a function of one's VNC session. If everyone used this, there would be no conflicts. However, not everyone uses this. -p rand will randomly pick a port range. This sometimes works. The probability decreases as the number of developers on a workstation increases. -p <number> will use a port the user specifies. If the user knows the port range is not in use, then this will typically be successful.

It is inconvenient to find a port range that is not in use. One can use the shell command "netstat | grep localhost | sort" to see what ports are in use but the data are not organized in an easy-to-use way. One can further play 20 questions, using the shell command "netstat | grep localhost | grep "\:nnn" where nnn is the first three digits of a possible port range. If this comes up empty for nnn and nnn+1, then one can typically be successful. But this can involve a fair amount of guesswork.

It would be nice to have a script that would find appropriate values of nnn from the "netstat | grep localhost" output.

It would be even nicer if this could be integrated with install_local_hadoop as a new -p option (say, -p auto). But for the moment, we'll satisfy ourselves with just creating the script.

> Need a better way to pick ports for install_local_hadoop
> --------------------------------------------------------
>
>                 Key: TRAFODION-1591
>                 URL: https://issues.apache.org/jira/browse/TRAFODION-1591
>             Project: Apache Trafodion
>          Issue Type: Improvement
>          Components: sql-general
>    Affects Versions: 2.0-incubating
>            Reporter: David Wayne Birdsall
>            Assignee: David Wayne Birdsall
>            Priority: Minor
>




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