Diamond is being deprecated, I checked for remaining collectors used in Cloud VPS:
[x] role::labs::instance: diamond::collector { 'SSHSessions' }
** This is a custom collector deployed via Puppet (modules/diamond/files/collector/sshsessions.py), it should be pretty straightforward to convert. I think this would even be a useful metric to be collected in general (i.e. also for production), integration could be done via the prometheus-node-exporter textfile collector (see prometheus::node_intel_microcode).
[x] role::labs::instance: diamond::collector diamond::collector::minimalpuppetagent { 'minimal-puppet-agent' }
** Replaced with prometheus::node_puppet_agent
[x] class puppetmaster::gitsync (used by role::puppetmaster::standalone): diamond::collector { 'CherryPickCounter' }
** This is a custom collector deployed via Puppet (modules/puppetmaster/files/cherry-pick-counter-collector.py). it should be pretty straightforward to convert.
[x] toollabs::services: diamond::collector { 'SGE' } (custom exporter from puppet.git)
** Will need to be ported to a custom exporter (if still needed).
[x] profile::toolforge::services::basic: diamond::collector { 'SGE':
** Will need to be ported to a custom exporter (if still needed).
[x] profile::toolforge::redis: diamond::collector { 'Redis' }
** In production metrics were ported to Prometheus in T148637, that can probably be reused/adapted.
** NOTE: The redis collector setup in prod requires PuppetDB to discover the nodes to scrape
[] define labstore::nfs_mount: diamond::collector { 'Nfsiostat' }
** This is a custom collector deployed via Puppet (modules/diamond/files/collector/nfsiostat.py), ~~I doubt there's an existing Prometheus exporter for this use case, so probably the existing one needs to be converted.~~ Can likely be replaced with the `mountstats` module of node_exporter (T210993#5769021)
[] toollabs::mailrelay: diamond::collector::extendedexim { 'extended_exim_collector': }; profile::toolforge::mailrelay: diamond::collector::extendedexim { 'extended_exim_collector': }
** In production metrics were ported to Prometheus in T179565, that can probably be reused/adapted.
** NOTE: The exim mtail collector setup in prod requires PuppetDB to discover the nodes to scrape
[x] class toollabs: diamond::collector::localcrontab { 'localcrontabcollector': }; profile::toolforge::grid::base: diamond::collector::localcrontab { 'localcrontabcollector': }
** Needs some investigation for replacement.
[] role::beta::availability_collector: diamond::collector { 'VarnishStatus' }
** This is a custom collector deployed via Puppet (modules/diamond/files/collector/varnishstatus.py), it should be pretty straightforward to convert (per watroles it's currently in use on deployment-cache-upload04, so seems still in use).