Load balancing in a storage system
Techniques are used for balancing load on a storage system according to multiple variables. The techniques may be used to provide, among other things, defining, across at least two variables, a balance constraint for a load on a storage system. Among a set of transfers of volumes from one node to another node in the storage system, a transfer of a volume that minimizes the distance between the load and an ideal balanced state of the storage system is identified. The identified transfer of a volume is added to a combination of transfers of volumes. Whether the combination of transfers of volumes meets the balance constraint is determined. If the combination meets the balance constraint, the combination is selected as a solution to balance the load.
1. A method comprising:
defining, across at least two variables, a balance constraint for a load on a storage system;
identifying, among a set of transfers of volumes from one node to another node in the storage system, a transfer of a volume that minimizes the distance between the load and an ideal balanced state of the storage system;
adding the identified transfer of a volume to a first combination of transfers of volumes;
determining whether the first combination of transfers of volumes meets the balance constraint;
if the first combination meets the balance constraint, selecting the first combination as a first potential solution to balance the load;
determining a set of combinations of transfers of volumes from one node to another node in the storage system;
applying, to each combination in the set, a fitness function that is based on the balance constraint;
selecting, from the set of combinations of transfers of volumes, a second combination as a second potential solution to balance the load, based on scores output by the fitness function;
selecting, from the first and second potential solutions, a solution to balance the load of the storage system; and
transferring volumes according to the solution.
2. The method of claim 1 , wherein selecting the solution comprises:
selecting the solution based on computing resources required by the first and second potential solutions.
3. The method of claim 2 , wherein the computing resources include input/output operations, processor cycles, total size of the volumes transferred in a combination, number of volumes transferred, and any combination thereof.
4. A method comprising:
defining, across at least two variables, a balance constraint for a load on a storage system;
identifying, among a set of transfers of volumes from one node to another node in the storage system, a transfer of a volume that minimizes a distance between the load and an ideal balanced state of the storage system;
adding the identified transfer of a volume to a combination of transfers of volumes;
determining whether the combination of transfers of volumes meets the balance constraint;
if the combination meets the balance constraint, selecting the combination as a solution to balance the load; and
transferring volumes according to the solution.
5. The method of claim 4 , wherein the distance between the load and the ideal balanced state of the storage system is a Euclidean distance or a Manhattan distance.
6. The method of claim 4 , further comprising:
repeating the steps of identifying transfers of volumes that minimize the distance between the load and an ideal balanced state of the storage system and adding identified transfers of volumes to the combination until the combination meets the balance constraint.
7. The method of claim 6 , further comprising:
after finding the combination that meets the balance constraint, continuing to identify transfers of volumes that minimize the distance between the load and an ideal balanced state of the storage system and adding identified transfers of volumes to the combination until no further transfers of volumes reduce the distance.
8. The method of claim 4 , further comprising:
repeating the steps of identifying transfers of volumes that minimize the distance between the load and an ideal balanced state of the storage system and adding identified transfers of volumes to a combination a predetermined number of times.
9. A method comprising:
defining, across at least two variables, a balance constraint for a load on a storage system;
determining combinations of transfers of volumes from one node to another node in the storage system;
applying, to each combination of transfers of volumes, a fitness function based on the balance constraint;
selecting a combination of transfers of volumes based on a score from the fitness function, as a solution to balance the load; and
transferring volumes according to the solution.
10. The method of claim 9 , wherein the fitness function evaluates a distance between a load adjusted by the combination of transfers of volumes and an ideal balanced state of the storage system.
11. The method of claim 9 , wherein the fitness function evaluates a number of transfers of volumes in a given combination of transfers of volumes.