Mechanism for a work node scan process to facilitate cluster scaling
A method and system that perform cluster node work assignments for tenants in a multi-tenant system. The method includes selecting a partition of a cluster tenant tracking structure having a hierarchical namespace and virtual nodes representing each tenant by a cluster node foreman controller, searching the selected partition for a tenant that is unlocked in the cluster tenant tracking structure, and initiating work for the tenant by the cluster node.
1. A method of cluster node work assignment for tenants in a multi-tenant system, the method comprising:
selecting, by a foreman controller of a cluster node, a partition of a cluster tenant tracking structure having a hierarchical namespace and virtual nodes representing each tenant of the multi-tenant system;
searching the selected partition for a tenant of the multi-tenant system that is unlocked in the cluster tenant tracking structure; and
initiating work for the unlocked tenant by the cluster node.
2. The method of cluster node work assignment of claim 1 , wherein selecting the partition is in response to expiration of a timer for the cluster node, or an increase in work capacity of the cluster node.
3. The method of cluster node work assignment of claim 1 , wherein the cluster tenant tracking structure is a tree of the virtual nodes with the hierarchical namespace, and the partition is a sub-tree of the tree.
4. The method of cluster node work assignment of claim 3 , wherein the virtual nodes indicate whether a tenant is assigned to the cluster node.
5. The method of cluster node work assignment of claim 1 , further comprising:
receiving notification of a tenant joining a cluster of the cluster node by a cluster node manager.
6. The method of cluster node work assignment of claim 5 , further comprising:
assigning the tenant to a partition of the cluster tenant tracking structure by the cluster node manager.
7. The method of cluster node work assignment of claim 1 , wherein selecting the partition for cluster tenant tracking structure is by round robin.
8. A non-transitory machine-readable storage medium that provides instructions that, if executed by a processor, will cause said processor to perform operations comprising:
selecting, by a foreman of a cluster node, a partition of a cluster tenant tracking structure having a hierarchical namespace and virtual nodes representing each tenant of a plurality of tenants;
searching the selected partition for a tenant, of the plurality of tenants, that is unlocked in the cluster tenant tracking structure; and
initiating work for the unlocked tenant by the cluster node.
9. The non-transitory machine-readable storage medium of claim 8 , wherein selecting the partition is in response to expiration of a timer for the cluster node, or an increase in work capacity of the cluster node.
10. The non-transitory machine-readable storage medium of claim 8 , wherein the cluster tenant tracking structure is a tree of virtual nodes with the hierarchical namespace, and the partition is a sub-tree of the tree.
11. The non-transitory machine-readable storage medium of claim 10 , wherein the virtual nodes indicate whether a tenant is assigned to the cluster node.
12. The non-transitory machine-readable storage medium of claim 8 , the operations further comprising:
receiving notification of a tenant joining a cluster of the cluster node by a cluster node manager.
13. The non-transitory machine-readable storage medium of claim 12 , the operations further comprising:
assigning the tenant to a partition of the cluster tenant tracking structure by the cluster node manager.
14. The non-transitory machine-readable storage medium of claim 8 , wherein selecting the partition for cluster tenant tracking structure is by round robin.
15. A computing device in a multi-tenant environment, the computing device comprising:
a non-transitory machine-readable medium having stored therein a foreman controller of a cluster node; and
a processor coupled to the non-transitory machine-readable medium, the processor to execute the cluster node and the foreman controller, the foreman controller to select a partition of a cluster tenant tracking structure having a hierarchical namespace and virtual nodes representing each tenant of the multi-tenant environment, search the selected partition for a tenant of the multi-tenant environment that is unlocked in the cluster tenant tracking structure, and initiate work for the unlocked tenant by the cluster node.
16. The computing device of claim 15 , wherein selecting the partition is in response to expiration of a timer for the cluster node, or an increase in work capacity of the cluster node.
17. The computing device of claim 15 , wherein the cluster tenant tracking structure is a tree of virtual nodes with the hierarchical namespace, and the partition is a sub-tree of the tree.
18. The computing device of claim 17 , wherein the virtual nodes indicate whether a tenant is assigned to the cluster node.
19. The computing device of claim 15 , further comprising:
a cluster node manager to receive notification of a tenant joining a cluster of the cluster node.
20. The computing device of claim 19 , further comprising:
a cluster node manager to assign the tenant to a partition of the cluster tenant tracking structure by the cluster node manager.
21. The computing device of claim 15 , wherein selecting the partition for cluster tenant tracking structure is by round robin.