Monitoring performance of a logically-partitioned computer
View Patent ↗An apparatus, system, and storage medium that in an embodiment collect a performance metric of a first partition in a logically-partitioned computer. If the difference between the performance metric and an expected performance metric exceeds a threshold, then a job or another partition is shut down or suspended. The expected performance metric is calculated based on the performance that is expected if the first partition is the only partition.
1. A non-transitory storage medium encoded with instructions, wherein the instructions when executed comprise:
determining whether a computer is partitioned into a plurality of logical partitions;
if the computer is partitioned into the plurality of logical partitions, collecting a plurality of first performance metrics in a sampling interval for the plurality of respective logical partitions of the computer and collecting a plurality of second performance metrics for a plurality of respective jobs in a first partition of the plurality of logical partitions in the computer, wherein the plurality of first performance metrics and the plurality of second performance metrics are not collected if the computer is not partitioned into the plurality of logical partitions;
reporting a plurality of differences between the plurality of first performance metrics for the plurality of respective logical partitions and a plurality of expected performance metrics for the plurality of respective logical partitions, wherein the plurality of expected performance metrics are expected to occur when the respective partition is an only partition executing in the computer, wherein the reporting the plurality of differences between the plurality of first performance metrics for the plurality of respective logical partitions and the plurality of expected performance metrics for the plurality of respective logical partitions further comprises displaying data on a video display;
calculating a plurality of degradations between the plurality of respective second performance metrics for the plurality of respective jobs and a plurality of anticipated performance metrics for the plurality of respective jobs in the first partition, wherein the plurality of anticipated performance metrics are expected to occur when no other jobs are executing, and wherein the plurality of anticipated performance metrics are associated with a plurality of respective types of the plurality of respective jobs;
summing the plurality of degradations into a total degradation for the plurality of the jobs in the first partition;
reporting the total degradation for the plurality of the jobs in the first partition;
determining whether the total degradation for the plurality of the jobs in the first partition exceeds a threshold; and
if the total degradation for the plurality of the jobs in the first partition exceeds the threshold, shutting down a second partition of the plurality of logical partitions in the computer.
2. The non-transitory storage medium of claim 1 , wherein the plurality of first performance metrics comprise an average number of cycles per instruction for the respective partition during the sampling interval.
3. The non-transitory storage medium of claim 1 , wherein the plurality of first performance metrics comprise processor utilization for the respective partition during the sampling interval.
4. The non-transitory storage medium of claim 1 , wherein the determining whether the total degradation for the plurality of the jobs in the first partition exceeds the threshold is periodically performed.
5. The non-transitory storage medium of claim 1 , wherein the plurality of first performance metrics comprise disk arm utilization.
6. A computer comprising:
a processor; and
memory connected to the processor, wherein the memory is encoded with instructions, wherein the instructions when executed on the processor comprise:
determining whether the computer is partitioned into a plurality of logical partitions,
if the computer is partitioned into the plurality of logical partitions, collecting a plurality of first performance metrics in a sampling interval for the plurality of respective logical partitions of the computer and collecting a plurality of second performance metrics for a plurality of respective jobs in a first partition of the plurality of logical partitions in the computer, wherein the plurality of first performance metrics and the plurality of second performance metrics are not collected if the computer is not partitioned into the plurality of logical partitions,
reporting a plurality of differences between the plurality of first performance metrics for the plurality of respective logical partitions and a plurality of expected performance metrics for the plurality of respective logical partitions, wherein the plurality of expected performance metrics are expected to occur when the respective partition is an only partition executing in the computer, wherein the reporting the plurality of differences between the plurality of first performance metrics for the plurality of respective logical partitions and the plurality of expected performance metrics for the plurality of respective logical partitions further comprises displaying data on a video display,
calculating a plurality of degradations between the plurality of respective second performance metrics for the plurality of respective jobs and a plurality of anticipated performance metrics for the plurality of respective jobs in the first partition, wherein the plurality of anticipated performance metrics are expected to occur when no other jobs are executing, and wherein the plurality of anticipated performance metrics are associated with a plurality of respective types of the plurality of respective jobs,
summing the plurality of degradations into a total degradation for the plurality of the jobs in the first partition,
reporting the total degradation for the plurality of the jobs in the first partition,
determining whether the total degradation for the plurality of the jobs in the first partition exceeds a threshold, and
if the total degradation for the plurality of the jobs in the first partition exceeds the threshold, shutting down a second partition of the plurality of logical partitions in the computer.
7. The computer of claim 6 , wherein the plurality of first performance metrics comprise an average number of cycles per instruction for the respective partition during the sampling interval.
8. The computer of claim 6 , wherein the plurality of first performance metrics comprise processor utilization for the respective partition during the sampling interval.
9. The computer of claim 6 , wherein the determining whether the total degradation for the plurality of the jobs in the first partition exceeds the threshold is periodically performed.
10. The computer of claim 6 , wherein the plurality of first performance metrics comprise disk arm utilization.