Hello all,
We are migrating to a new ArcGIS Enterprise setup and I am looking for some advice or thoughts on the following (generalized) data strategy that I am looking to implement:
- A branch versioned geodatabase for internal authoritative datasets: required for Roads & Highways and other versioned data edited via services.
- A traditional versioned geodatabase for internal authoritative datasets: we will need some direct feature class editing that is not available with branch versioning. Some internal datasets won't need versioning and will simply b appended to.
- A traditional versioned geodatabase for incoming third party datasets to keep separate from the internal authoritative datasets. Versioning will not be enabled on these as they will never be edited, just overwritten. Question below if this one is required or are hosted services a better option and use a file geodatabase instead.
AWS cloud environment with Postgres for databases.
For the third party datasets, which can range in size from a few MB to well over 10GB, and will only ever be read-only and overwritten when an update comes our way, what is the better approach; a referenced service or a hosted service in Portal? Keeping it general that feature services are required and only a small set of these will also use Map Image or Vector Tiling when required.
If the better approach is a hosted service, is storing these datasets in an enterprise geodatabase necessary and offer any benefits over using a file geodatabase?
We are moving away from everyone having access to the SDE connections to the service centric approach with only admins having access to SDE. Most of our services are added to ArcGIS Online as referenced services with Portal used as the admin center.
I know there will be special case here and there, but there is generally nothing complex we do with or use the data for outside of some branch editing, complete overwrite, or appending.
Thanks for any thoughts / advice.
All the best,
Glen
