[Draft] Proposal - Incorporate meshsync as library and provide an option to run meshsery server with meshsync inside to get a cluster snapshot on start

[Nikita B]

Hey, hello everybody   !

As by Lee's suggestion. sending this message to the mailing list (same as in #meshery slack channel).

Following one of the discussions about allowing users to very quickly try meshery with their cluster resources, but without requirement for deploying anything additional like CRDs, meshsync, meshery operator, meshsery broker, I have prepared a brief potential proposal on this topic:
https://docs.google.com/document/d/1iANgnZT5i6OdtxI__CTHs3XLHz5oZ9Ly3vp-DZ_D3OA/edit?usp=sharingFeel free to leave a comment there 
Thank you 

[Lee Calcote]

Hi Nikita,

This is most excellent. Given our conversations on the subject, I took the liberty of (largely) rewriting the proposal, which I'm hopeful is 1) most helpful and 2) most welcome.

In addition, I dropped a new item in the "Unresolved Questions" section in consideration of whether the proposed embedding of MeshSync is boot time or a connection time concern. Answering this question is probably the foremost consideration to make before advancing the spec any further.

Despite my positioning of embedded MeshSync as an all-in-one deployment model, I will argue that it is easier for us to implement and more flexible for the user to have this functionality exposed as a configurable option of every individual Kubernetes’ connection.

- Lee