Document how to deploy BETYdb on the terraref network

[dlebauer]

• [x] Add the following about deploying BETYdb

---

Deploying BETYdb

• Option 1 (required skill: programmer or ambitious scientist): install on a server
• Option 2 (required skill: scientist who can use bash) install the PEcAn Virtual Machine

For option 2 (the PEcAn VM), need to replace the database

• VM has data that is not relevant to TERRAREF
• VM has default users and passwords (carya / illinois). At a minimum, these must be changed.
• VM has a machine id of 99. Any new data would be deleted by load.bety.sh that imports data from other instances of BETYdb (like terraref's BETYdb number 6)

Importing TERRA REF data

(this does not cover sharing data, only importing from the TERRA REF public reference dataset. Your data is safe there are four steps required for your data to be shared on the network: 1) mark data as access_level=4. 2) run dump.db.sh 3) put the dump on a public url. 4. add the dump url to betydb.org/machines)

• If you want to join the network and import data from the TERRA REF team, here are some considerations:
• The BETYdb that is provided with the PEcAn virtual machine contains data outside of the scope of TERRA
• Machines on the network need their own unique sync ID 'X'. All new records created on this machine will have row numbers starting with X billion. Any edits to data with numbers under 1 billion or in the 6 billions will be deleted on sync.
• To import the the public terraref data use the load.bety.sh with MYSITE=X and REMOTESITE = 6 as described in the BETYdb technical documentation (bety --> docs --> under "Distributed BETYdb").
  • this will populate your database with metadata related to the sorghum genotypes, trait definitions, plot and management definitions for trials in AZ, and protocols (methods), etc. If you set up a cron job for load.bety.sh all of our data will be inserted into your database as it is released.
  • As other TERRA teams get sync ID Y and want to share data, load.bety.sh can be run with REMOTESITE = Y (e.g. REF is 6, MEPP is 8, TAMU has 9, Purdue 11, Clemson 12).
• To stay in sync, you also need to periodically update the software. It is fairly easy to do. Sign up for alerts of new releases here https://lists.illinois.edu/lists/subscribe/betydb-releases

[dlebauer]

now there are lots of examples of how to deploy using docker