Reorganization of a data set
A computer-implemented method (CIM), according to one embodiment, includes performing a reorganization analysis for mitigating latency during reorganization of a data set. The reorganization analysis includes using regression analysis to predict a time duration that reorganization of the data set will take based on a total size of the data set, relating the predicted time duration to a previous data set reorganization time duration, and performing linear extrapolation to estimate a time at which the total size of the data set will exceed a predetermined percentage of a maximum size of the data set. The method further includes reorganizing the data set before the estimated time, where the data set is reorganized at a time determined based on the reorganization analysis or derivative thereof.
1. A computer-implemented method (CIM), the CIM comprising:
performing a reorganization analysis for mitigating latency during reorganization of a data set, the reorganization analysis including:
using regression analysis to predict a time duration that reorganization of the data set will take based on a total size of the data set,
relating the predicted time duration to a previous data set reorganization time duration, and
performing linear extrapolation to estimate a time at which the total size of the data set will exceed a predetermined percentage of a maximum size of the data set; and
reorganizing the data set before the estimated time, wherein the data set is reorganized at a time determined based on the reorganization analysis or derivative thereof.
2. The CIM of claim 1 , wherein the regression analysis used to predict the time duration that the reorganization of the data set will take includes linear regression analysis.
3. The CIM of claim 1 , wherein the reorganization analysis includes: building a time series model to forecast activity of the data set, wherein building the time series model includes: monitoring read and write activity recorded in a journal data set on a direct access storage device (DASD) by using historical read and write activity data.
4. The CIM of claim 3 , wherein the data set is a collection of metadata.
5. The CIM of claim 3 , wherein the reorganization analysis includes: estimating an amount of time until the data set reaches the maximum size of the data set, wherein the estimating is performed by extrapolating, from the regression analysis, utilization of the data set over time.
6. The CIM of claim 5 , wherein the reorganization analysis includes: forecasting from the time series model to determine, over the amount of time until the data set reaches the maximum size of the data set, periods of relatively low utilization of the data set, finding a local minima of the periods of relatively low utilization, and determining whether activity during the local minima of the periods of relatively low utilization has at least a predetermined degree of similarity with the historical read and write activity data over a same relative period of time.
7. The CIM of claim 6 , wherein the time determined based on the reorganization analysis or derivative thereof is the local minima of the periods of relatively low utilization.
8. The CIM of claim 7 , wherein the time determined based on the reorganization analysis or derivative thereof is the local minima of the periods of relatively low utilization in response to a determination that the activity during the local minima of the periods of relatively low utilization has at least a predetermined degree of similarity with the historical read and write activity data over a same relative period of time.
9. A computer program product (CPP), the computer program product comprising:
a set of one or more computer-readable storage media; and
program instructions, collectively stored in the set of one or more storage media, for causing a processor set to perform the following computer operations:
perform a reorganization analysis for mitigating latency during reorganization of a data set, the reorganization analysis including:
using regression analysis to predict a time duration that reorganization of the data set will take based on a total size of the data set,
relating the predicted time duration to a previous data set reorganization time duration, and
performing linear extrapolation to estimate a time at which the total size of the data set will exceed a predetermined percentage of a maximum size of the data set; and
reorganize the data set before the estimated time, wherein the data set is reorganized at a time determined based on the reorganization analysis or derivative thereof.
10. The CPP of claim 9 , wherein the regression analysis used to predict the time duration that the reorganization of the data set will take includes linear regression analysis.
11. The CPP of claim 9 , wherein the reorganization analysis includes: building a time series model to forecast activity of the data set, wherein building the time series model includes: monitoring read and write activity recorded in a journal data set on a direct access storage device (DASD) by using historical read and write activity data.
12. The CPP of claim 11 , wherein the data set is a collection of metadata.
13. The CPP of claim 11 , wherein the reorganization analysis includes: estimating an amount of time until the data set reaches the maximum size of the data set, wherein the estimating is performed by extrapolating, from the regression analysis, utilization of the data set over time.
14. The CPP of claim 13 , wherein the reorganization analysis includes: forecasting from the time series model to determine, over the amount of time until the data set reaches the maximum size of the data set, periods of relatively low utilization of the data set, finding a local minima of the periods of relatively low utilization, and determining whether activity during the local minima of the periods of relatively low utilization has at least a predetermined degree of similarity with the historical read and write activity data over a same relative period of time.
15. The CPP of claim 14 , wherein the time determined based on the reorganization analysis or derivative thereof is the local minima of the periods of relatively low utilization.
16. The CPP of claim 15 , wherein the time determined based on the reorganization analysis or derivative thereof is the local minima of the periods of relatively low utilization in response to a determination that the activity during the local minima of the periods of relatively low utilization has at least a predetermined degree of similarity with the historical read and write activity data over a same relative period of time.
17. A computer system (CS), the CS comprising:
a processor set;
a set of one or more computer-readable storage media; and
program instructions, collectively stored in the set of one or more storage media, for causing the processor set to perform the following computer operations:
perform a reorganization analysis for mitigating latency during reorganization of a data set, the reorganization analysis including:
using regression analysis to predict a time duration that reorganization of the data set will take based on a total size of the data set,
relating the predicted time duration to a previous data set reorganization time duration, and
performing linear extrapolation to estimate a time at which the total size of the data set will exceed a predetermined percentage of a maximum size of the data set; and
reorganize the data set before the estimated time, wherein the data set is reorganized at a time determined based on the reorganization analysis or derivative thereof.
18. The CS of claim 17 , wherein the regression analysis used to predict the time duration that the reorganization of the data set will take includes linear regression analysis.
19. The CS of claim 17 , wherein the reorganization analysis includes: building a time series model to forecast activity of the data set, wherein building the time series model includes: monitoring read and write activity recorded in a journal data set on a direct access storage device (DASD) by using historical read and write activity data.
20. The CS of claim 19 , wherein the data set is a collection of metadata.