Hi Pravin,
Glad to know you are interested ! Great progress so far.
Sachin, can you review the code in more detail?
My first thoughts are :
1. Please write a blog post detailing your understanding of Fedora metrics
ecosystem and statscache, how you progressed, any obstacles you faced and
how you debugged them. Documenting your work is as important as writing
code as it enables community to get to know about your work, learn about
different parts of the project and helps in onboarding newcomers.
2. Please comment the code. It's pretty clean, but a few comments will
surely be more helpful for newcomers.
3. Work on your proposal - both pdf and fedora wiki page. Please review the
proposal and especially the timeline with mentors. This is crucial. Start
soon on this as reviews take time.
4. If you haven't already, don't forget to send an introduction to
summer-coding mailing list of fedora.
5. Optional, since you are familiar with statscache, please update
documentation of statscache wherever you feel is necessary. Any issues you
close or commits you make to repo count positively towards your proposal.
Hope this helps,
Best,
Bee
On Sun, Apr 2, 2017 at 10:39 PM, Pravin Chaudhary <
pravin.chaudhary.me(a)gmail.com> wrote:
Hello!
I am Pravin Chaudhary, an undergrad from India.
I would like to participate in GSoC 17 with the project Centralized
Metrics generation.
<
https://fedoraproject.org/wiki/Summer_coding_ideas_for_2017>
Initially a friend of mine(vashramc) wrote an email asking for the details
in this project. The requirements leading up to the proposal were mentioned
as:
*As an initial contribution, we would love to see a few of these things -
if not all,1. An understanding of Fedora metrics ecosystem 2. An
understanding of statscache code3. Try deploying a local instance of
statscache4. Optional, Build a basic plugin in statscache for eg. which
just records how many messages are emitted by fedmsg daily.*
I have done all of the above; I have perused Statscache code in detail and
that of the plugins as well. I have a clear idea of how the system works
overall and also of the individual components.
As required, I have written a plug-in for statscache
<
https://github.com/pravinkc/statscache_plugins/pull/1> that shows all
the references to a list of users. The list of users we wish to track is
defied via the environment variable TRACK_USERS, providing it comma
separated usernames.
I will draft my proposal, once I receive comments on my progress so far.
Thank you!
Pravin