Create Sync Mechanism Between Two GeoNodes (used to be Proposal)

[cgiovando]

Related to to #12, we need a proposal to develop and implement the following two options:

• Mono-directional sync between the local (editable) and the online instances (read-only). In this scenario, changes are only allowed in the local instance, and pushed asynchronously over a limited-capacity internet connection to the online instance.

• Bi-directional sync between the local and online instances, where changes are allowed on both. Syncing happen asynchronously over the same limited-capacity internet connection. This scenario requires options to manage versions and conflicts.

[simboss]

Hi Cristiano this is what driving the proposal we are putting together. Feel free to comment on it to refine it.

It is something more than -1- and something less than -2- in your comment above and I would stick with it for the moment.

[gmolinario]

hi @simboss and @cgiovando I took a look at the document, thanks for starting that.

One major comment at this stage:

Could we include in the document two or three scenarios:

1) What you outlined already in the document, a bidirectional master/slave syncing with full functionality (as you said, choosing also which content needs to be synced etc. ), 2) a simpler bidirectional "mirror" sync where there are two copies of the GeoNode in different network locations that are kept in completel mirror sync (master-master?) and 3) a scenario in which the sync is only monodirectional - aka the master is a local geonode for example at a ministry in afghanistan, and the "slave" is on the internet. The master has full functionality and the slave has read-only functionality.

The goal is to understand what are the differences in costs associated with the development, and be able to propose to the TTL consequently.

By all means add to the discussion if you think some of the above scenarios are NOT useful at all, or you think there are other ones that are more useful and should be proposed.

[simboss]

@gmolinario a few points below.

1. My proposal is NOT bidirectional, is unidirectional with the possibility to have local content not synched. So I would say it 3 plus something.

2. Developing a bidirectional mechanism that is reliable will require a couple of weeks of design + a few months of development. If I had to shoot a ROM budget I would shoot 40k$.

Let me know your thoughts.

[gmolinario]

hi @simboss thanks for the clarification. I added a comment to the above directly in the proposal google doc. My goal here is to have something I can present that has a few different options (your #1 above being your preferred one). Please @cgiovando if you can take a look at the proposal and comment it would be great.

[gmolinario]

I am re-opening this issue. It would be good to plan out timeline and next steps.