Hadoop-ecosystem ports

[Kurt Jaeger]

Hi!

> I will be working on making Hadoop a first class citizen on FreeBSD.

Thanks, that sounds helpful, as that is a large group of applications.

> This will include bumping the existing ports for Hadoop, HBase, Oozie,
> Spark (probably some others as well), and creating ports for HUE, Sqoop,
> Zeppelin, Flume, Ambari and a variety of other projects maintained by
> the ASF. Think Hortonworks HDP/Cloudera CDH/MapR, just on FreeBSD.
> 
> This can't be done on just one laptop  and I'm looking for people
> interested in testing and reviewing them.

> Preferably also someone with hardware to stand up a large enough cluster
> as reference architecture, to show that Hadoop on FreeBSD trumps Hadoop
> on Linux.

I can provide a 32 GB / 4c+4t box as a testing environment, if it
helps.

> What's a good mailing list to coordinate this effort on? Given that
> they're mostly Apache projects, would apache@ be a good fit?

apache@ looks very limited to the webserver (if you review
the last few month of posts).

For a start, use ports at . If ports@ gets too much hadoop-related
posts, you can always ask postmaster@ to add a list, like
freebsd-hadoop@ or something similar.

https://wiki.freebsd.org/Teams/postmaster#How_to_create_a_new_list.3F

says:

What we need

    Terse (one-line) description of list.
    More verbose description of list.
    Any deviations from these defaults:
        Max message body size: 200KB
        Not moderated
        Open (subscription not required for) posting
        Public archives
        "Official" name of list "foo" is "freebsd-foo"
        List is a technical list (vs. freebsd-chat@)
        List is for discussions (vs. freebsd-announce@)
        No designated non-postmaster "moderator" or "list administrator" 

-- 
pi at opsec.eu            +49 171 3101372                    2 years to go !