Merge pull request #3 from vimc/i2270

I2270 - update docs and diagram

.gitignore

@@ -1,3 +1,4 @@
 staging/disk
 staging/shared/vault_config
 .vagrant
+.idea

docs/Annex.md

@@ -6,6 +6,8 @@ Host key fingerprint (ECDSA): `SHA256:lYJUME+1ATyTvzcFvs8K+aSOBfX/Xv3mpz3kKeYdzM
 ## Services
 * Annex database (in docker container)
 * Barman
+* Barman cached metrics exporter
+* bb8 metrics exporter
 * Cron job to nightly dump data from barman and ferry it to the starport
 * Cron job to nightly back up starport to AWS using duplicati
 * (Starport - this is just a folder than can be SSH'd to by bb8 from other machines)
@@ -17,12 +19,12 @@ sudo apt upgrade
 sudo reboot now
 
 sudo su montagu
-docker start barman-montagu
-docker exec barman-montagu setup-barman --no-initial-backup
+cd montagu-db-backup && ./start-metrics.sh
+cd montagu-bb8/bb8 && ./starport-metrics/run /home/montagu/starport
 
 ```
-Annex database will resume automatically. Cron job will continue running as per
-usual. Our goal is to have Barman resuming automatically also, but as yet this does not work. See https://vimc.myjetbrains.com/youtrack/issue/VIMC-2116
+Annex database and barman will resume automatically. Cron job will continue running as per
+usual.
 
 We couldn't upgrade because VirtualBox was running. If so, you may find this
 untested script that should hibernate all VirtualBox VMs that are running. You
@@ -37,3 +39,5 @@ vboxmanage list runningvms \
 ### Check things are back up and running
 * Annex database: `docker exec -it db_annex psql -U vimc -d montagu`
 * Barman (as user `montagu`): `barman-montagu status`
+* Barman-metrics: `curl localhost:5000/metrics`
+* bb8-metrics: `curl localhost:5001/metrics`