Rebased. Use /usr/local/etc/periodic. Clean certbot_dir.

add hostname

DNS: grafana. CNAME www-dev.nasqueron.org

Deployment can be using sqlite3 as long as it's still performant
as we want our monitoring tools to be resiliant.

follow comments

Probably a good part of roles/core/monitoring when grains["os_family"] == "RedHat". Eglide has "Debian" for that grain, but not sure if we've enough RAM there.

chmod of files and folder + encryption

Just for reference, this was a test deployment. This is not currently installed on Dwellers, and needs to be in Salt as part of T650.

This task has been created in 2016 to publish metrics from PCP (Performance Co-Pilot) on RHEL-like servers, especially our Docker engines.

RabbitMQ exporters have been added to NetBox under the tag observability -> https://netbox.nasqueron.org/ipam/services/?tag=observability 🔒

Next: memcached

Pending container redeployment with D3374, we can reach metrics set in D3373 with socat:

+port

Deployed on both Docker engines, but docker-002 is still to be restarted, probably good idea to sync that with dnf update

Also, we need to declare Dovecot ports at https://netbox.nasqueron.org/virtualization/virtual-machines/10/ services table (on the public IP)

I suspect the version 1.6.1 (currently in packages) is compatible with FreeBSD 13 while the version 1.8.2 is compatible with FreeBSD 14.

Align names

hunt-insecable-spaces

Avoid duplicate jinja context for service config

hunt-insecable-spaces

Configure systemd service to restrict to private IP too

no-op on FreeBSD

Service name is different on EPEL package

Successfully deployed on FreeBSD servers:

Independently of the 2020 plan for a service mesh, we're going to deploy Prometheus right now to gain observability on currently deployed services.

Prometheus has gained in traction and timeseries become now more and more standard.