Supun,
Yup, you've caught on a bit to what I was getting at.
The monitor, as Eric hinted as, grew out of necessity for debugging and
development purposes of Accumulo. Functionality is typically added from
a "it would be nice to know X" stance. Along this line, this is how
ACCUMULO-594 came about.
Thinking about optimizing the time we contribute to Accumulo, not having
to write our own monitoring presentation would, theoretically, free up
our time to work on issues inside of Accumulo. This is where the "kill
the monitor" ticket comes from. Personally, creating functionality to
hook up Accumulo to other presentation systems supplements what the
monitor already shows, and it would take many Accumulo releases to
generate enough confidence in a complete replacement that would merit
remove of the existing monitor.
Now, from your stance, it's up to you to decide what interests you most.
Don't be afraid that contributions made to the Accumulo monitor will be
invalidated. Any additions that Eric mentioned (not to mention many more
that we haven't thought of) would be immediately useful for users. If
you're familiar with other pieces of monitoring software and would like
to work on connecting it to Accumulo, that'd be great too!
Perhaps we can move towards what you'd like to get out of this
experience and try to tailor the goals to meet that?
On 3/26/13 11:35 AM, Supun Kamburugamuva wrote:
Hi Josh,
Thank you for the prompt response.
As Eric mentioned in ACCUMULO-594 I can see there are lot of things
that can be done in the monitoring space. Some of the things include
improving JMX. Also in one Jira it says to kind of abandon the monitor
and use another library (monitoring tool) for monitoring using JMX.
Also there are Jiras for improving the existing monitor visualization
as well. (so I'm bit confused)
As you've mentioned I think it would be really helpful to define the scope.
Thanks,
Supun..
On Tue, Mar 26, 2013 at 11:17 AM, Josh Elser <[email protected]> wrote:
Supun,
I'd be happy to help out in this regard.
I've been trying to catch up on what I need to do to be an "official"
mentor, but we can use the dev list or the ACCUMULO-594 ticket to start
documenting how and what you want to approach the project.
I know the ticket was rather open-ended; perhaps we can work towards
refining its scope a bit more? Start working towards getting you and all of
us on the same page.
- Josh
On 3/26/13 10:39 AM, Supun Kamburugamuva wrote:
Hi all,
I would like to learn and contribute to projects in big data space and
would really like to do a GSoc project in Accumulo. I'm a PHD student
at Indiana University and my research area is in Distributed Systems.
I'm also a committer to few Apache projects. I would like to take a
project in Accumulo and contribute. Please let me know if this is
possible.
Regards,
Supun..
