prometheus: increase retention duration to 2y

With the move to pluto the setup is now benefitting from ZSTD compression
yielding a compression factor of 2.14x, bringing our data down from 75GB
to a mere 35GB.

We also doubled the disk size to 512GB, so spending ~150GB on metrics is
not only feasible, but will allows us to compare performance metrics over
longer intervals.

Author: mweinelt

delft/pluto/prometheus/default.nix

@@ -41,7 +41,7 @@
   services.prometheus = {
     enable = true;
     extraFlags = [
-      "--storage.tsdb.retention=${toString (150 * 24)}h"
+      "--storage.tsdb.retention=${toString (720 * 24)}h"
       "--web.external-url=https://prometheus.nixos.org/"
     ];
     globalConfig.scrape_interval = "15s";