Managing heterogeneous cluster environment
AssigneesInternational Business Machines Corporation, University College Dublin
InventorsHitham Ahmed Assem Aly Salama, Teodora Sandra Buda, Patrick J. O'Sullivan, Philip Perry, Lei Xu
Life of this patent
Estimating the term…
Abstract
An embodiment includes a method for use in managing a cluster comprising a plurality of computers, each computer comprising at least one hardware processor coupled to at least one memory. The method comprises a computer-implemented manager: collecting access pattern data for the plurality of computers; based at least in part on the access pattern data, determining that at least a first one of the plurality of computers is overutilized; and responsive to determining that the first computer is overutilized, moving stored data from the first computer to at least a second one of the plurality of computers.
Claims19
A method for use in managing a cluster comprising a plurality of computers, each computer comprising at least one hardware processor coupled to at least one memory, the plurality of computers within the cluster comprising a leader node and a plurality of follower nodes, the method comprising:
- the leader node of the cluster collecting access pattern data for the plurality of follower nodes within the cluster;
- based at least in part on the access pattern data, the leader node of the cluster determining that at least a first one of the plurality of follower nodes within the cluster is overutilized by analyzing a historical access pattern;
- and
- responsive to determining that the first follower node is overutilized, selecting a subset of data stored at the first follower node to move based on a determination that the subset of data is frequently accessed and the leader node of the cluster moving the selected subset of data stored on the first follower node to at least a second one of the plurality of follower nodes within the cluster to configure the first follower node and the second follower node to be subjected to similar access patterns.
A leader node within a cluster of computers, the cluster of computers further comprising a plurality of follower nodes, each follower node comprising at least one hardware processor coupled to at least one memory, the leader node comprising:
- a memory;
- and
- at least one processor coupled with the memory, the processor operative:
- to collect access pattern data for the plurality of follower nodes within the cluster;
- based at least in part on the access pattern data, to determine that at least a first one of the plurality of follower nodes within the cluster is overutilized by analyzing a historical access pattern;
- and
- responsive to determining that the first follower node is overutilized, selecting a subset of data stored at the first follower node to move based on a determination that the subset of data is frequently accessed and the leader node of the cluster moving the subset of data stored on the first follower node to at least a second one of the plurality of follower nodes within the cluster to configure the first follower node and the second follower node to be subjected to similar access patterns.
A cluster of computers comprising:
- a leader node comprising at least one hardware processor coupled to at least one memory;
- and
- a plurality of follower nodes, each follower node comprising at least one hardware processor coupled to at least one memory;
- the leader node of the cluster being operative to collect access pattern data for the plurality of follower nodes within the cluster;
- based at least in part on the access pattern data, the leader node of the cluster being operative to determine that at least a first one of the plurality of follower nodes within the cluster is overutilized by analyzing a historical access pattern;
- and
- responsive to determining that the first follower node is overutilized, selecting a subset of data stored at the first follower node to move based on a determination that the subset of data is frequently accessed and the leader node of the cluster moving the subset of data stored on the first follower node to at least a second one of the plurality of follower nodes within the cluster to configure the first follower node and the second follower node to be subjected to similar access patterns.
Dependencies are inferred from the claim text.