Dynamic buffer allocation in similar infrastructures

Patent metadata

Field	Value
Publication number	US-10735342-B2
Application number	US-201916692037-A
Country	US
Kind code	B2
Filing date	Nov 22, 2019
Priority date	Oct 4, 2017
Publication date	Aug 4, 2020
Grant date	Aug 4, 2020

Abstract

A method, computer program product, and system includes a processor(s) obtaining historical data related to buffer space usage of a common shared storage resource by a group of journals. The processor(s) monitors each journal over repeating temporal periods and determined a predictability for each journal (i.e., predictable or unpredictable. The processor(s) generates usage statistics for each journal during each monitored period dependent on whether the journal was predictable or unpredictable during the monitored period. For each monitored period, the processor(s) ranks predictable journals by a buffer space requirement for each journal of the predictable journals. Based on the rankings, the processor(s) determined a buffer space requirement for each journal for a current period, where the current period shares temporal qualities with the given monitored period. The processor(s) allocates, for use during the current period, based on the buffer space requirement, buffer space to each journal.

First claim

What is claimed is:

1. A computer program product comprising:
a computer readable storage medium readable by one or more processors and storing instructions for execution by the one or more processors for performing a method comprising:
obtaining, by one or more processors, historical data related to buffer space usage for each data structure in a group of data structures comprising a computing system, wherein each data structure utilizes buffer space comprising a shared storage resource, by monitoring activity in each data structure over repeating temporal periods;
determining, by the one or more processors, for each data structure, based on the obtained historical data, a predictability for each data structure for each monitored period of the repeating temporal periods, wherein the predictability is selected from the group consisting of: predictable and unpredictable, wherein a predictable data structure, for a monitored period, uses a predictable amount of buffer space during the monitored period, and an unpredictable data structure, for the monitored period, uses an unpredictable amount of buffer space during the monitored period;
generating, by the one or more processors, based on the historical data, usage statistics for each data structure during each monitored period, wherein the usage statistics generated for each monitored period are dependent on whether the data structure was predictable or unpredictable during the monitored period;
determining, by the one or more processors, a buffer space requirement for each data structure of the group of data structures for a current period, wherein the current period shares one or more temporal qualities with a given monitored period; and
based on the determining, allocating, by the one or more processors, for use during the current period, based on the buffer space requirement for each data structure, buffer space to each data structure of the group of data structures, from the buffer space in the shared storage resource.

2. The computer program product of claim 1 , wherein determining the buffer space requirement for each data structure of the group of data structures for the current period comprises:
determining, by the one or more processors, a buffer size threshold for each data structure based on the predictability of each data structure for the given monitored period;
determining, by the one or more processors, a portion of the buffer space in the shared storage resource allocated to one or more data structures of the predictable data structures in the given monitored period, available to reclaim for allocating to one or more data structures of the unpredictable data structures in the given monitored period, the use during the current period; and
setting, by the one or more processors, a buffer size and a threshold buffer size for each data structure.

3. The computer program product of claim 2 , the method further comprising:
in advance of or during a second period, obtaining, by the one or more processors, an event related to one or more data structures;
determining, by the one or more processors, that the event that will breach the threshold buffer size of the one or more data structures of the group of data structures during the second period;
determining, by the one or more processors, additional buffer space requirements for the one or more data structures, based on the event; and
automatically re-allocating, by the one or more processors, a portion of buffer space in the shared storage resource to the one or more data structures, based on predictability of the one or more data structures during a second monitored period, wherein the second monitored period and the second period share one or more temporal qualities, the additional space requirements, and the portion of the buffer space available to reclaim.

4. The computer program product of claim 3 , further comprising:
designating, by one or more programs, the one or more data structures previously determined to be predictable for the second monitored periods as unpredictable for the second monitored period.

5. The computer program product of claim 3 , wherein automatically re-allocating the portion of buffer space further comprises:
determining, by the one or more processors, an additional buffer space requirement for each data structure for the second period, based on the rate of change of actual buffer usage of each data structure.

6. The computer program product of claim 3 , the method further comprising:
monitoring, by the one or more processors, a number of times the one or more processors automatically re-allocate buffer space to each of the one or more data structures, based on obtaining events related to the one or more data structures;
based on the number exceeding a first threshold, designating, by the one or more processors, the one or more data structures as rogue; and
compressing, by the one or more processors, messages stored in the one or more data structures.

7. The computer program product of claim 6 , the method further comprising:
based on the number exceeding a second threshold, rejecting, by the one or more processors, all further incoming messages destined for the one or more data structures.

8. The c computer program product of claim 7 , wherein the second threshold exceeds the first threshold.

9. The computer program product of claim 8 , the method further comprising:
notifying, by the one or more processors, an administrator that the one or more data structures were designated rogue.

10. The computer program product of claim 1 , wherein generating the usage statistics for each data structure during each monitored period comprises:
determining, by the one or more processors, for predictable data structures in the group of data structures, a minimum usage, a maximum usage, an average usage, and a standard deviation and for unpredictable data structures in the group of data structures, an average predicted buffer usage and a standard deviation from the average predicted buffer usage.

11. The computer program product of claim 1 , wherein allocating buffer space to each data structure of the group of data structures further comprises:
determining, by the one or more processors, a combined buffer space requirement for the group of data structures;
determining, by the one or more processors, that buffer space comprising the shared storage resource is less than combined buffer space requirement; and
notifying, by the one or more processors, an administrator to provision additional shared storage space for the computing system.

12. The computer program product of claim 1 , wherein determining the buffer space requirement for each data structure of the group of data structures for a current period, is based on rankings by the one or more processors, for the given monitored period, wherein the ranking comprises:
for each monitored period, ranking, by the one or more processors, predictable data structures for the monitored period, by the buffer space requirement for each data structure of the predictable data structures.

13. The computer program product of claim 12 , the method further comprising:
based on the rankings for a second given monitored period, for a period subsequent to the current period, determining, by the one or more processors, an updated buffer space requirement for each data structure of the group of data structures for the period subsequent to the current period, wherein the period subsequent to the current period shares one or more temporal qualities with the second given monitored period; and
based on the determining, adjusting, by the one or more processors, for use during

Assignees

• IBM

Inventors

• Kairali Sudheesh S
• Kalangumvathakkal Vijay R
• Kumar Jagdish
• Singh Sunil Kumar S

Classifications

• G06F3/0604 Primary

  Improving or facilitating administration, e.g. storage management · CPC title

• H04L47/83

  based on usage prediction · CPC title

• G06F3/0631

  by allocating resources to storage systems · CPC title

• G06F3/067

  Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS] · CPC title

• H04L47/72 Primary

  using reservation actions during connection setup · CPC title