The project proposal i submitted (removed the personal information, academic and work/internship Experience portion)

*Project Proposal*

There would be two significant portions to this project whose ultimate aim is to build upon the existing GeoDns platform that Mozilla.org already has to provide better load balancing to the infrastructure.

The first part would be to include the function of weighted resource record. With this functionality, the DNS server would be able to spread the load over the mirrors in proportion of the weight configured for each mirror.

Modification to the way the DNS answer the query would also need to be done so that the percentage of time a mirror is used is close to the weight configured for the resource. The implementation of the “per resource weight” would probably include a change in database schema so as to store the weight for each resource. Another modification to the database would be to store the usage statistics for each resource. These data would allow the reporting feature to be implemented in the management tool to be built in the second part.

The second part of this project would be to build an improved management tool for this platform. The improved management tool should be able to perform the configuration of the new weighted resource record feature implemented in the first part of the project. The tool should also be capable of displaying the percentage of times each resource in the geographical reason had been used.

The results for each phase will be released in incremental bits so that the work can be reviewed by the mentors and the community. With input from more people would make the project more beneficial the Mozilla community.

I had previous had experience with DNS deployment and IP QOS configuration. My knowlegde in the working of the DNS protocol and different method of implementation would be useful in this project. Since I had not much experience with GSLB, i had started reading up on some common implmentation of this setup.

Success Criteria
-   The platform should be able to store the per resource weight record
-   The platform would return the correct answer so that the load is spread according to the weight configured
-   A management tool should be built to able the resource weight record to be configured easily
-   The end results should be useful for the Mozilla community

*Schedule of Deliverables*

I would divide the project into different stages:

Stage 1 [Until end of May]:
This stage would primarily be focused on the internal workings of the platform. Specifically the logic of choosing the ideal resource that should be used for answering DNS queries. The general area that requires addition of code and functions should also be identified.
Deliverables: Documentation of proposed changes to the system.

Stage 2 [1st to 31th June]:
The objective of this stage is to have the weight of resource record stored in the database and modify the DNS server to return the correct resource as the answer. Deliverables: Database schema for storing the weights for each resource. Patches made to DNS server for selecting and return answer to query according to the “per resource weight”.

Stage 3 [1st to 31st July]: The building of the management tool would be done in this stage. Work will involve testing of the tool to ensure that the tool functions as describe in the design specifications.
Deliverables: Management tool and usage notes.

Stage 4 [Until mid-August]: The final stage of this project would be to clean up the code added and documentation kept over the project. Time would also be invested on doing the final testing to take note of whatever improvement can be done in the future.
Deliverables: Project change and implementation documentation.

Other commitments: I will be taking 1 module at my univeristy from 11 May to 20 June. This will take up about 4 hours each week. However, I do not foresee it affecting the deliverables timeline as I manage my time quite well.

Why Mozilla*
I chose Mozilla Project because I feel that it is made up of a very active and enthusiastic community. I am also impressed with the community because of its ability to produce quality software like Firefox and Thunderbird which had shaped the browser market significantly. The other reason would be because there are not any other organization offering projects in the building of infrastructure. After my experiences in different field of IT, it is things like service deployment and server management that interest me most. I feel that the experience in coding of the service component would allow me to appreciate things better.

I had changed the status or the 3 proposals that i previously posted to private. Still considering if the proposals actually give out too much personal information to the public. and this blog had been getting spam comments. should install captcha to the comments soon.