List of projects others might want to work on

[stack <st...@duboce.net>]

Recently on IRC, there was a request for a list of projects/areas that 
noobie hbase developers could dig-in on.  I'm posting one such list here 
in case others were wondering the same:

1. InfoServers' UI extension/revamp: Expose more server attributes and 
counts, lists of regions by table, graphing of regionserver loading 
history, etc
2. Help with mapfile rewrite: Experiment with 
algorithms/format/compressions to improve the fundamental hbase building 
block
3. Write a PIG loader for hbase and PIG LATIN scripts to do bulk table 
upload and backup
4. Improve hbase performance/latency (core dev is currently focused on 
robustness and scaling).
5. Build an hbsck tool, a tool to run common repairs to an hbase 
install. Develop referential integrity to support this tool (Tool could 
be written in java or even in groovy TBD).
6. Write clone of google app engine python datastore module

Ask here (or on IRC) if any questions about the above.
Thanks,
St.Ack

Re: List of projects others might want to work on

["Jim R. Wilson" <wi...@gmail.com>]

7. Support for TS in Mutation object to allow atomic row mutations
with different timestamps (needed for atomic clone, which is needed
for simplest possible rowkey alter ability)
8. Single jar with simplest possible hadoop/hbase/rest/thrift ready to
run on common architectures (for people who want to just try it out -
the 10 minute test)
9. phpHbaseAdmin - like phpMyAdmin, but for hbase ;) (requires the
ability to retrieve timestamp data from cells - rumored to be
available in 0.2)
10. APT packages for hbase/hadoop (apt-get install hbase)
11. Lucene/bloom filter layer on top of put/get/delete to allow
constantly updated search indexes (admittedly, a wild idea)

-- Jim R. Wilson (jimbojw)

On Thu, Apr 10, 2008 at 2:37 PM, stack <st...@duboce.net> wrote:
> Recently on IRC, there was a request for a list of projects/areas that
> noobie hbase developers could dig-in on.  I'm posting one such list here in
> case others were wondering the same:
>
>  1. InfoServers' UI extension/revamp: Expose more server attributes and
> counts, lists of regions by table, graphing of regionserver loading history,
> etc
>  2. Help with mapfile rewrite: Experiment with
> algorithms/format/compressions to improve the fundamental hbase building
> block
>  3. Write a PIG loader for hbase and PIG LATIN scripts to do bulk table
> upload and backup
>  4. Improve hbase performance/latency (core dev is currently focused on
> robustness and scaling).
>  5. Build an hbsck tool, a tool to run common repairs to an hbase install.
> Develop referential integrity to support this tool (Tool could be written in
> java or even in groovy TBD).
>  6. Write clone of google app engine python datastore module
>
>  Ask here (or on IRC) if any questions about the above.
>  Thanks,
>  St.Ack
>