ARK2/Avalon2

From ARK
Revision as of 21:16, 15 December 2016 by John Layt (talk | contribs)

Jump to: navigation, search

Avalon 2

Avalon is the project management ARK for LP Archaeology. Avalon triggers the creation of a job folder on the local server, which is then synced to the other servers. If a GIS is required, then standard shapefiles are created in the job folder, and all the GIS data is stored in the job folder. Any Project that also requires a site recording ARK then sets that up separately.

A number of issues exist here:

  • File syncing is slow, resource intensive, and error prone
  • GIS data is heavily duplicated, wasting space
  • GIS data is fractured, i.e. finding what data we already have requires manual searching of folders, and may lead to the purchase of duplicate data
  • Use of ARK for data recording is time-consuming to set up so often not done
  • Many jobs are deemed too small to bother entering the data in ARK, this makes querying our data corpus or automating many report writing functions impossible
  • Versioning files, submitting them for review, and checking they have been is a manual process and open to abuse

ARK 2 provides several new features that can alleviate this:

  • File management and file versioning support
  • Workflow management
  • Easy automated creation of new ARK instances
  • Multiple sites and schemas in a single ARK install
  • Support for running on PostgreSQL, and consequently support for PostGIS

The proposal for Avalon 2 is as follows:

  • A single ARK install running both Avalon and all site ARKs, either as a single instance or two instances sharing a single user database.
  • The database will run on PostgreSQL, providing both the Avalon/ARK data server and a PostGIS database in a single install.
  • The GIS database will manage as many shared vector and raster resources as possible, as well as the GIS for each Project.
  • On creation of a new Project, the PM can choose to create a GIS and an ARK, or trigger this at any time later.
  • A connection to Harvest will be created using their API, allowing for combined reporting, and automated creation of projects and allocation of staff.
  • Later Avalon could seek to replicate the required features of Harvest itself.