out of shared memory in Waku fleets database nodes

[Ivansete-status]

We currently have wrongly configured the shm_size in the database docker containers, e.g.: node-db-01.do-ams3.waku.sandbox

Maybe we could establish a minimum of '1gb', and not 0 as it is currently.

[jakubgs]

Fixed:

• infra-waku#4f202094 - node-db: fix shared memory calculation
• infra-status#f869ae11 - node-db: fix shared memory calculation

We were casting 0.3 to int which made it 0. Order of operations matters.