staging cluster is in rather high memory pressure

@gsamson noticed it on late wednesday.

The cluster management node (rancher-node-staging-rke2-mgmt1) has a high load, and high memory usage [1]. I'll increase its memory.

Furthermore, the overall cluster is in warning memory level for some time. So, I'll add one more node to the staging cluster to try to ease that part too.

[1] https://grafana.softwareheritage.org/d/QvtaMQyWk/node-exporter-full?orgId=1&refresh=1m&var-DS_PROMETHEUS=Thanos&var-job=node&var-node=rancher-node-staging-rke2-mgmt1.internal.staging.swh.network&var-diskdevices=%5Ba-z%5D%2B%7Cnvme%5B0-9%5D%2Bn%5B0-9%5D%2B%7Cmmcblk%5B0-9%5D%2B

[2]

Refs. #4876 (closed)

#4699 (closed) may help too.