Optimal storage and workload placement, and high resiliency, in geo-distributed cluster systems
Technologies for cluster systems that are natively geo-site-aware. Such a cluster system makes use of this awareness to determine the subsets of nodes located at various geo-sites at physical configuration, to optimize workload placement based on the geo-sites, to make failover and failback decisions based on the geo-sites, and to assign voting and prune nodes for quorum management based on the geo-sites. Such capabilities result in cluster systems that are more resilient and more efficient in terms of resource usage than cluster systems without such native geo-site awareness.
1. A method performed on a cluster system comprising at least one computing device that includes at least one processor and memory, the method comprising:
automatically identifying, by the cluster system, a plurality of geo-site objects, each of the geo-site objects corresponding to a geographical location and identifying a subset of physical resources of the cluster system that is physically located at the corresponding geographical location;
determining a client geographical location at which a client is physically located;
placing, by the cluster system and based on the determination, a workload of the client on a workload node of a first geo-site object of the plurality of geo-site objects, where the first geo-site object identifies a subset of physical resources of the cluster system that is physically located at the client geographical location;
placing, by the cluster system, dependent storage of the workload on a storage node of the geo-site object, where the workload node and the storage node are cluster nodes of the cluster system; and
pruning, based on a node not being in the geo-site object, the node from one of the plurality of geo-site objects other than the geo-site object.
2. The method of claim 1 where the storage node and the workload node are the same cluster node.
3. The method of claim 1 further comprising placing, upon a failover involving the workload, the workload on a third node of the geo-site object.
4. The method of claim 3 further comprising placing, in response to a failback event involving the workload subsequent to the placing the workload on the third node, the workload back on the workload node of the geo-site object.
5. The method of claim 1 further comprising identifying the geo-site object as representing a preferred geo-site of the cluster system.
6. The method of claim 1 where the workload node is identified by the cluster system as a preferred node of the cluster system.
7. A cluster system comprising:
a workload node that comprises at least one processor and memory coupled to the at least one processor, the memory comprising computer-executable instructions that, based on execution by the at least one processor, configure the cluster system to:
identify a plurality of geo-site objects, each of the geo-site objects corresponding to a geographical location and identifying a subset of physical resources of the cluster system that is physically located at a corresponding geographical location of the cluster system;
determine a client geographical location at which a client is physically located;
place, based on the determination, a workload of the client on the workload node of a first geo-site object of the plurality of geo-site objects, where the first geo-site object identifies a subset of physical resources of the cluster system that is physically located at the client geographical location; and
place dependent storage of the workload on a storage node of the geo-site object, where the workload node and the storage node are cluster nodes of the cluster system; and
a geo-site quorum mange via which the cluster system is configured to prune, based on a node not being in the geo-site object, the node from one of the plurality of geo-site objects other than the geo-site object.
8. The cluster system of claim 7 where the storage node and the workload node are the same cluster node.
9. The cluster system of claim 7 further configured to place, upon a failover involving the workload, the workload on a third node of the geo-site object.
10. The cluster system of claim 9 further configured to place, in response to a failback event involving the workload subsequent to the placing the workload on the third node, the workload back on the workload node of the geo-site object.
11. The cluster system of claim 7 further configured to identify the geo-site object as representing a preferred geo-site of the cluster system.
12. The cluster system of claim 7 where the workload node is identified by the cluster system as a preferred node of the cluster system.
13. At least one computer-readable medium that comprises computer-executable instructions that, based on execution by at least one computing device of a cluster system, configure the cluster system to perform actions comprising:
automatically identifying, by the cluster system, a plurality of geo-site objects, each of the geo-site objects corresponding to a geographical location and identifying a subset of physical resources of the cluster system that is physically located at the corresponding geographical location;
determining a client geographical location at which a client is physically located;
placing, by the cluster system and based on the determination, a workload of the client on a workload node of a first geo-site object of the plurality of geo-site objects, where the first geo-site object identifies a subset of physical resources of the cluster that is physically located at the corresponding geographical location;
placing, by the cluster system, dependent storage of the workload on a storage node of the geo-site object, where the workload node and the storage node are cluster nodes of the cluster system; and
pruning, based on a node not being in the geo-site object, the node from one of the plurality of geo-site objects other than the geo-site object.
14. The at least one computer-readable medium of claim 13 where the storage node and the workload node are the same cluster node.
15. The at least one computer-readable medium of claim 13 , the actions further comprising placing, upon a failover involving the workload, the workload on a third node of the geo-site object.
16. The at least one computer-readable medium of claim 15 , the actions further comprising placing, in response to a failback event involving the workload subsequent to the placing the workload on the third node, the workload back on the workload node of the geo-site object.
17. The at least one computer-readable medium of claim 13 where the geo-site object is identified by the cluster system as representing a preferred geo-site of the cluster system, or where the workload node is identified by the cluster system as a preferred node of the cluster system.