Hi Shivang,
Thanks a lot for going through the proposal.
You're right, I haven't explained how I'm going to resolve the MongoDB issues.
I tried to adhere to the 10-20 sentence limit given for the project description
in the Template so I couldn't fit everything in the proposal.
I don't have a solution for it already but I'll tell you whatever ideas I've
had in my mind.
The first idea is crude but it can potentially be effective. I could create an
initialization/shutdown sequence where mongodb --repair is used routinely.
(https://docs.mongodb.com/manual/reference/program/mongod/#cmdoption-mongod-repair)
The mongodb --repair command fixes many common issues relating to incorrect
shutdown methods resulting in corruption and it would at least solve the more
commonly occurring trivial issues.
The second course of action is investigating occurrences of this issue at
existing deployments and trying to find the cause of that. I will need to
communicate with the deployment admins and the IIAB community members who have
faced these issues for the exact errors and the sequence of actions that led to
these errors.
The third thing is that I've deployed Sugarizer on two Raspberry Pi's that are
always running and a VPS. Maybe with some load testing and some automation
testing (login, user creation and other tasks that have db writes), I will be
able to find more such mongoDB issues.
So with a combination of these three methods, I should be able to solve
numerous mongoDB stability issues. I had planned to utilize the Community
Bonding period for more detailed discussion and guidance from the community and
my potential mentors. I will start thinking about a more specific solution for
this problem.
Thanks,
Rishabh Nambiar.
‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
On March 25, 2018 7:15 PM, Shivang Shekhar <[email protected]> wrote:
> Hi Rishabh,
> I went through your proposal and looks really good about how you have
> arranged things to be done.
>
> The only thing that I felt missing in the proposal was the contingency case
> or how exactly you are going to test the db crashes (in your case I suppose
> its mongo).
> Though you have discussed the problem but do you have a plan in mind how to
> resolve one or some fair idea about how things can be sorted as this would be
> the biggest issue when it comes to scaling the project.
>
> also,
> ping me the line number if you think that you have this covered in your
> proposal and I was somewhat not able to catch it.
>
> Cheers
>
> On Fri, Mar 23, 2018 at 8:20 AM Rishabh Nambiar <[email protected]>
> wrote:
>
>> Hello, this is Rishabh.
>> I know I've already shared my proposal with all of you but we've been asked
>> to share the Google Docs version of our proposals with the mentors for
>> easier feedback and review.
>>
>> Here's the link:
>> https://docs.google.com/document/d/1g_3IVCw1g0ghT4s8Cdg8AlkFyalQIYiC7K1QE08iy7o/edit?usp=sharing
>>
>> Feedback of any kind will be valuable, Thank you!
>>
>> Regards,
>> Rishabh Nambiar
>
>
> https://mailtrack.io/ Sent with
> [Mailtrack](https://chrome.google.com/webstore/detail/mailtrack-for-gmail-inbox/ndnaehgpjlnokgebbaldlmgkapkpjkkb?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality)
> --
>
> Shivang Shekhar
>
> Mentor - Google Summer of Code 2018
>
> Co-Chair (WebVR Industry Committee) - VRAR Association
_______________________________________________
Sugar-devel mailing list
[email protected]
http://lists.sugarlabs.org/listinfo/sugar-devel
