Comparison
Cell-based architecturevsMulti-tenancy
Cell-based architecture
customers are divided across eight complete copies of the stack, so a bad deploy or a poison request takes out an eighth of them.
Partitioning the whole system into independent, complete instances and assigning each customer to one. It converts most total outages into partial ones and gives deployments a natural containment unit. The costs are operational multiplication — everything you do, you do eight times — and routing, since something has to map a customer to a cell and that router is now the shared dependency.
Full entry →Multi-tenancy
several teams or customers share one cluster, and every performance question becomes 'is it us or is it them'.
Running multiple independent workloads or customers on shared infrastructure. It is what makes a platform economical, and it introduces isolation, fairness, quota and noisy-neighbour problems that dedicated infrastructure simply does not have. The critical distinction is whether the tenants trust one another: sharing between internal teams and sharing between hostile customers require entirely different boundaries.
Full entry →