US 12,327,020 B2
Volume replication of stateful-sets
Padmanabha Venkatagiri Seshadri, Mysore (IN); and Krishnasuri Narayanam, Bangalore (IN)
Assigned to INTERNATIONAL BUSINESS MACHINES CORPORATION, Armonk, NY (US)
Filed by INTERNATIONAL BUSINESS MACHINES CORPORATION, Armonk, NY (US)
Filed on Dec. 28, 2022, as Appl. No. 18/147,692.
Prior Publication US 2024/0220114 A1, Jul. 4, 2024
Int. Cl. G06F 3/06 (2006.01)
CPC G06F 3/0611 (2013.01) [G06F 3/0635 (2013.01); G06F 3/065 (2013.01); G06F 3/067 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computing device comprising:
a processor;
a storage device coupled to the processor;
a code stored in the storage device, wherein an execution of the code by the processor configures the computing device to perform a method, comprising:
receiving a plurality of input requests and corresponding plurality of input request types;
for each container of a plurality of containers, extracting a plurality of control flow paths and mapping the plurality of input request types to the plurality of control flow paths;
learning control flow path-to-volume mappings by computing, for each control flow path of the plurality of control flow paths, (i) one or more associations between the control flow path and one or more volumes and (ii) a nature of access of the one or more volumes;
for each pod replica, computing a dominant control flow path by computing a most common control flow path and a most common volume corresponding to the most common control flow path, wherein the dominant control path comprises:
labeling edges in the plurality of control flow paths, obtained from a mapping of the plurality of input request types to plurality of control flow paths, with volume traffic and volume identification information;
generating a cluster of paths and a cluster of volumes; and
selecting the most common control flow path from the cluster of paths and the most common volume corresponding to the selected most common control flow path; and
responsive to receiving a request for volume management of an identified most common volume, reducing a latency by locks on shared artifacts by determining whether a volume replication is warranted for each pod replica, and based on the determination either:
replicating the identified most common volume based on volume management information comprising information about the nature of access of the identified most common volume and/or information about a locking pressure of the identified most common volume, or
creating one or more views of the identified most common volume for pod sharing.