5 ms·
You can have a distributed system that is partially available too. Consider a GPU cluster. A network partition occurs. Both nodes can suspend jobs or refuse
by throwawaymaths 3y ago
You can have a distributed system that is partially available too. Consider a GPU cluster. A network partition occurs. Both nodes can suspend jobs or refuse to accept jobs that require more than one nodes worth of compute until the partition is healed, while still independently serving jobs that require only one node, while healing all accounting details when reconciliation is possible.