Long-range distributed resource planning using workload modeling in hyperconverged computing clusters
Systems for computing cluster management. One embodiment commences upon receiving a set of observed workload parameters corresponding to one or more observable workloads that run in a computing cluster. While the workloads are running, workload stimulus and cluster response observations are taken and used to generate a workload resource usage predictive model based on mappings or correlations between the observable workloads parameters and observed resource usage measurements. A set of planned workloads are applied to the workload resource usage predictive model to predict a set of corresponding predicted resource usage demands. The predicted resource usage demands are then mapped to a set of recommended hardware to form resource deployment recommendations that satisfy at least some of the corresponding resource usage demands while also observing a set of hardware model compatibility constraints. The resource deployment recommendations that satisfy the set of hardware model compatibility constraints are displayed in a user interface.
1. A method comprising:
receiving a first set of workload parameters corresponding to a first workload executing on a first node of a first type in a first cluster;
receiving a second set of workload parameters corresponding to a second workload executing on a second node of the first type in a second cluster;
generating a single predictive model based at least on both the first set of workload parameters associated with the first cluster and the second set of workload parameters associated with the second cluster;
emitting a first resource deployment recommendation and a second resource deployment recommendation based at least in part on resource usage predictions generated by the single predictive model;
wherein the second resource deployment recommendation is different from the first resource deployment recommendation; and
displaying the first resource deployment recommendation or the second resource deployment recommendation.
2. The method of claim 1 , wherein the first resource deployment recommendation or the second resource deployment recommendation is based at least in part on a hardware model compatibility.
3. The method of claim 1 , wherein the first resource deployment recommendation or the second resource deployment recommendation is based at least in part on one or more resource allocation rules comprising a replication factor, a compression factor.
4. The method of claim 3 , wherein the single predictive model maps a set of planned workload parameters to a set of predicted resource usage characteristics.
5. The method of claim 3 , wherein the single predictive model is associated with nodes of the first type.
6. The method of claim 1 , wherein the displaying of the first resource deployment recommendation or the second resource deployment recommendation includes displaying a time-based series of deployment events.
7. The method of claim 6 , wherein the single predictive model is one of a plurality of predictive models associated with respective types of nodes.
8. The method of claim 1 , wherein the displaying the first resource deployment recommendation or the second resource deployment recommendation further comprises displaying a target date.
9. The method of claim 1 , wherein the first resource deployment recommendation comprises a node of a first type, and the second resource deployment recommendation comprises a node of a second type.
10. The method of claim 1 , wherein the first resource deployment recommendation and the second resource deployment recommendation are generated by at least mapping at least some of the resource usage predictions generated by the single predictive model to a plurality of node types.
11. A non-transitory computer readable medium having stored thereon a sequence of instructions which, when executed by a processor performs a set of acts comprising:
receiving a first set of workload parameters corresponding to a first workload executing on a first node of a first type in a first cluster;
receiving a second set of workload parameters corresponding to a second workload executing on a second node of the first type in a second cluster;
generating a single predictive model based at least on both the first set of workload parameters associated with the first cluster and the second set of workload parameters associated with the second cluster;
emitting a first resource deployment recommendation and a second resource deployment recommendation based at least in part on resource usage predictions generated by the single predictive model,
wherein the second resource deployment recommendation is different from the first resource deployment recommendation; and
displaying the first resource deployment recommendation or the second resource deployment recommendation.
12. The computer readable medium of claim 11 , wherein the first resource deployment recommendation or the second resource deployment recommendation is based at least in part on a hardware model compatibility.
13. The computer readable medium of claim 11 , wherein the first resource deployment recommendation or the second resource deployment recommendation is based at least in part on one or more resource allocation rules comprising a replication factor, a compression factor.
14. The computer readable medium of claim 13 , wherein the single predictive model maps a set of planned workload parameters to a set of predicted resource usage characteristics.
15. The computer readable medium of claim 13 , wherein the single predictive model is associated with nodes of the first type.
16. The computer readable medium of claim 11 , wherein the displaying of the first resource deployment recommendation or the second resource deployment recommendation includes displaying a time-based series of deployment events.
17. The computer readable medium of claim 16 , wherein the single predictive model is one of a plurality of predictive models associated with respective types of nodes.
18. The computer readable medium of claim 11 , wherein the displaying the first resource deployment recommendation or the second resource deployment recommendation further comprises displaying a target date.
19. A system comprising:
a storage medium having stored thereon a sequence of instructions; and
processor that executes the sequence of instructions to cause a set of acts comprising:
receiving a first set of workload parameters corresponding to a first workload executing on a first node of a first type in a first cluster;
receiving a second set of workload parameters corresponding to a second workload executing on a second node of the first type in a second cluster;
generating a single predictive model based at least on both the first set of workload parameters associated with the first cluster and the second set of workload parameters associated with the second cluster;
emitting a first resource deployment recommendation and a second resource deployment recommendation based at least in part on resource usage predictions generated by the single predictive model,
wherein the second resource deployment recommendation is different from the first resource deployment recommendation; and
displaying the first resource deployment recommendation or the second resource deployment recommendation.
20. The system of claim 19 , wherein the first resource deployment recommendation or the second resource deployment recommendation is based at least in part on a hardware model compatibility.