Method and apparatus for load balancing virtual data movers between nodes of a storage cluster

1. A non-transitory tangible computer readable storage medium having stored thereon a computer program for implementing a method of load balancing virtual data movers (VDM) between nodes of a storage cluster, the computer program including a set of instructions which, when executed by a computer, cause the computer to perform a method comprising the steps of:

establishing a storage cluster including a plurality of nodes, one of the nodes implementing a cluster manager and each node having a system Virtual Data Mover (VDM) instantiated thereon;

assigning, by the cluster manager, primary responsibility for each of a plurality of data VDMs to corresponding nodes of the storage cluster, each data VDM having responsibility for at least one user file system;

assigning, by the cluster manager, corresponding backup nodes for the plurality of data VDMs;

collecting node statistics by each system VDM on each node in the cluster of nodes, the node statistics including operational parameters of the node and activity levels of the data VDMs on the node;

collecting the node statistics, by a cluster manager, from each of the system VDMs;

using the collected node statistics to assign a respective node score to each node in the storage cluster;

using the node scores to identify possible data VDM movement combinations within the storage cluster;

selecting one of the data VDM movement combinations for implementation within the storage cluster that will reduce disparity between node scores within the storage cluster; and

implementing the selected one of the data VDM movement combinations by moving at least some of the data VDMs between the nodes of the storage cluster.

2. The non-transitory tangible computer readable storage medium of claim 1 , further comprising weighting the node statistics in connection with using the collected node statistics to assign the respective node scores.

3. The non-transitory tangible computer readable storage medium of claim 2 , wherein the step of selecting one of the data VDM movement combinations for implementation comprises applying a first set of hard rules and a second set of soft rules.

4. The non-transitory tangible computer readable storage medium of claim 3 , wherein a first hard rule of the set of hard rules is that a given node's score must be above a high threshold for the that given node to be considered overloaded.

5. The non-transitory tangible computer readable storage medium of claim 3 , wherein a second hard rule of the set of hard rules is that a given node's score must be below a low threshold for that given node to be considered underloaded.

6. The non-transitory tangible computer readable storage medium of claim 3 , wherein a third hard rule of the set of hard rules is that a given node must have at least more than one data VDM assigned to it for the given node to be considered overloaded.

7. The non-transitory tangible computer readable storage medium of claim 3 , wherein a fourth hard rule of the set of hard rules is that a potential target node to receive a given data VDM in a data VDM movement process must not have a target node score that exceeds a node score of the node that is currently hosting the given data VDM.

8. The non-transitory tangible computer readable storage medium of claim 7 , wherein the target node score is a hypothetical node score for the target node based on the target node's current node score and an expected increase in node score attributable to movement of the given data VDM to the target node.

9. The non-transitory tangible computer readable storage medium of claim 3 , wherein the set of soft rules minimizes data VDM movement within a storage cluster.

10. The non-transitory tangible computer readable storage medium of claim 9 , wherein a first soft rule of the set of soft rules is that the data VDMs are assigned to nodes within the storage cluster, at an original time of provisioning the storage cluster to minimize subsequent movement of the data VDMs within the storage cluster after provisioning.

11. The non-transitory tangible computer readable storage medium of claim 9 , wherein a second soft rule of the set of soft rules is that data VDM movement of the data VDMs should be minimized by causing a given failover from the node hosting the given data VDM to the assigned corresponding backup node for the given data VDM rather than moving the given data VDM to a third node in the storage cluster.

12. The non-transitory tangible computer readable storage medium of claim 9 , wherein a third soft rule of the set of soft rules is that data VDM movement is proactively initiated in connection with other actions affecting a structure of the storage cluster.

13. The non-transitory tangible computer readable storage medium of claim 12 , wherein one of the other actions is addition of a new node to the storage cluster.

14. The non-transitory tangible computer readable storage medium of claim 12 , wherein one of the other actions is removal of one of the nodes from the storage cluster.

15. The non-transitory tangible computer readable storage medium of claim 1 , wherein the operational parameters of the node include:

a total number of data VDMs on the node; and

a number of user file systems on each data VDM on the node.

16. The non-transitory tangible computer readable storage medium of claim 1 , wherein the operational parameters of the node include:

front end statistics associated with high write pending conditions;

storage array events; and

memory usage levels.

17. The non-transitory tangible computer readable storage medium of claim 1 , wherein the activity levels of the data VDMs on the node include input/output activity levels of each file system on each data VDM on the node.