Increase prometheus retention duration to 3 years for all our clusters

[yuvipanda]

We initially set our prometheus retention to 1 year. However, since this is also where metrics about usage lie, we've gotten a few requests from communities (including one from cryocloud) to increase that.

Retention is currently set to 366d. Let's set it to 3 years (1098d) instead.

Definition of done

• [x] Retention config is set to 1098d in two places in the base support chart
• [x] Any retention overrides in specific clusters are discovered, and removed (so everyone gets an even 3y)
• [x] Prometheus servers are actively running on all clusters after this change is made
• [x] New issue is created to check on disk space usage on the prometheus servers a month later

[sgibson91]

Clusters checked for running prometheus servers:

• [x] 2i2c
• [x] 2i2c-aws-us
• [x] 2i2c-uk
• [x] awi-ciroh
• [x] catalystproject-africa
• [x] catalystproject-latam
• [x] cloudbank
• [x] earthscope
• [x] gridsst
• [x] hhmi
• [x] jupyter-health
• [x] jupyter-meets-the-earth
• [x] kitware
• [x] leap
• [x] linked-earth
• [x] nasa-cryo
• [x] nasa-esdis
• [x] nasa-ghg
• [x] nasa-veda
• [x] openscapes
• [x] opensci
• [x] pangeo-hubs
• [x] pchub
• [x] projectpythia
• [x] smithsonian
• [x] ubc-eoas
• [x] utoronto
• [ ] victor

[sgibson91]

New issue opened: https://github.com/2i2c-org/infrastructure/issues/4632