We are testing MCO in GCP using only 2 jobs instead of 3. Thereby, some jobs are failing reporting a 22 hours timeout.

Since we are adding new tests, those timeouts are happening more often. We need to modify the configuration in order to use 3 jobs before all GCP jobs start reporting timeouts in all the executions.

At least we should do it in 4.18 and 4.19. Likely, doing it in 4.17 would be a good idea too.

We should update the file https://docs.google.com/spreadsheets/d/1wSbcniu0tIwnx_VUrEL9u90jcHvzL9wRPv2UU_X4YPs/edit?gid=402564414#gid=402564414 when this ticket is closed.