Dell EMC PowerVault MD 34XX/38XX Series Storage Arrays Administrator's Guide

PDF

Interpreting performance monitor data

Performance Monitor provides you with data about devices. You can use this data to make storage array performance tuning decisions, as described in the following table:

Table 1. Performance data implications The following table describes the data to make storage array performance tuning decisions:
Performance Data	Implications for Performance Tuning
Total I/Os	This data is useful for monitoring the I/O activity of a specific RAID controller module and a specific virtual disk, which can help identify possible high-traffic I/O areas. You might notice a disparity in the total I/Os (workload) of RAID controller modules. For example, the workload of one RAID controller module is heavy or is increasing over time while that of the other RAID controller module is lighter or more stable. In this case, you might want to change the RAID controller module ownership of one or more virtual disks to the RAID controller module with the lighter workload. Use the virtual disk total I/O statistics to determine which virtual disks to move. You might want to monitor the workload across the storage array. Monitor the Total I/Os in the background performance monitor. If the workload continues to increase over time while application performance decreases, you might need to add additional storage arrays. By adding storage arrays to your enterprise, you can continue to meet application needs at an acceptable performance level.
IOs/sec	Factors that affect input/output operations per second (IOs/sec or IOPS) include these items: Access pattern (random or sequential) I/O size RAID level Cache block size Whether read caching is enabled Whether write caching is enabled Dynamic cache read prefetch Segment size The number of physical disks in the disk groups or storage array The transfer rates of the RAID controller module are determined by the application I/O size and the I/O rate. Generally, small application I/O requests result in a lower transfer rate but provide a faster I/O rate and shorter response time. With larger application I/O requests, higher throughput rates are possible. Understanding your typical application I/O patterns can help you determine the maximum I/O transfer rates for a specific storage array. You can see performance improvements caused by changing the segment size in the IOPS statistics for a virtual disk. Experiment to determine the optimal segment size, or use the file system size or database block size. For more information about segment size and performance, see the related topics listed at the end of this topic. The higher the cache hit rate, the higher I/O rates will be. Higher write I/O rates are experienced with write caching enabled compared to disabled. In deciding whether to enable write caching for an individual virtual disk, look at the current IOPS and the maximum IOPS. You should see higher rates for sequential I/O patterns than for random I/O patterns. Regardless of your I/O pattern, enable write caching to maximize the I/O rate and to shorten the application response time. For more information about read/write caching and performance, see the related topics listed at the end of this topic.
MBs/sec	See IOs/sec.
I/O Latency, ms	Latency is useful for monitoring the I/O activity of a specific physical disk and a specific virtual disk and can help you identify physical disks that are bottlenecks. Physical disk type and speed influence latency. With random I/O, faster spinning physical disks spend less time moving to and from different locations on the disk. Too few physical disks result in more queued commands and a greater period of time for the physical disk to process the command, increasing the general latency of the system. Larger I/Os have greater latency due to the additional time involved with transferring data. Higher latency might indicate that the I/O pattern is random in nature. Physical disks with random I/O will have greater latency than those with sequential streams. If a disk group is shared among several virtual disks, the individual virtual disks might need their own disk groups to improve the sequential performance of the physical disks and decrease latency. If inconsistency exists with physical disks of a common disk group. This condition might indicate a slow physical disk. With disk pools, larger latencies are introduced and uneven workloads might exist between physical disks making the latency values less meaningful and in general higher.
Cache Hit Percentage	A higher cache hit percentage is desirable for optimal application performance. A positive correlation exists between the cache hit percentage and the I/O rates. The cache hit percentage of all of the virtual disks might be low or trending downward. This trend might indicate inherent randomness in access patterns. In addition, at the storage array level or the RAID controller module level, this trend might indicate the need to install more RAID controller module cache memory if you do not have the maximum amount of memory installed. If an individual virtual disk is experiencing a low cache hit percentage, consider enabling dynamic cache read prefetch for that virtual disk. Dynamic cache read prefetch can increase the cache hit percentage for a sequential I/O workload.