eG Monitoring
 

Measures reported by OraZFSARCAStTest

The ARC is the Adaptive Replacement Cache and is an in-DRAM cache for file system and volume data. This test monitors the accesses to the ARC and reports performance issues in the ARC, if any. This test also helps to understand how well the current workload is caching in the ARC.

Outputs of the test: One set of results for the Oracle ZFS Storage System being monitored.

The measures made by this test are as follows:

Measurement Description Measurement Unit Interpretation
arcHitRatio Indicates the ARC hit ratio during the last measurement period. Percent

A high value is desired for this measure.

avgDataHits Indicates the average data hits during the last measurement period. Hits/sec

Data hit occurs when a data block is in the ARC DRAM cache and had returned.

A high value is desired for this measure. A very low value is a cause for concern.

peakDataHits Indicates the peak data hit during the last measurement period. Hits/sec

The detailed diagnosis shows when data requests had hit peaks.

avgDataMisses Indicates the average data misses during the last measurement period. Misses/sec

Data miss is when a data block is not in the ARC DRAM cache and is read from the L2ARC cache devices (if available and the data is cached on them) or the pool disks.

A high value is desired for this measure. A very low value is a cause for concern.

peakDataMisses Indicates the peak data misses during the last measurement period. Misses/sec

The detailed diagnosis shows when data requests had been unsuccessful thus causing degradation in performance.

avgMetaDataHits Indicates the average metadata hits during the last measurement period. Hits/sec

The term metadata refers to the “data about data” and is a concept utilized to store the data for all form modifications.

This measure displays the average count of successful metadata requests.

peakMetaDataHits Indicates the peak metadata hits during the last measurement period. Hits/sec

The detailed diagnosis shows when and how many metadata requests had been successful.

avgMetaDataMisses Indicates the average metadata misses during the last measurement period. Misses/sec

A high value of this measure can cause performance degradation of the system and increases latency.

peakMetaDataMisses Indicates the peak metadata misses during the last measurement period. Misses/sec

The detailed diagnosis shows when and how many metadata requests had been unsuccessful.

avgPrefetchedDataHits Indicates the average prefetched data hits during the last measurement period. Hits/sec

Prefetching is read-ahead caching. It lets the SP anticipate the data an application will request, so that it can read it from disk into its read cache before the data is needed.

A high value is desired for this measure.

peakPrefetchedDataHits Indicates the peak prefetched data hits during the last measurement period. Hits/sec

The detailed diagnosis shows when and how many prefetched data requests had been successful.

avgPrefetchedDataMisses Indicates the average prefetched data misses during the last measurement period. Misses/sec

A high value of this measure can cause performance degradation of the system and increases latency.

peakPrefetchedDataMisses Indicates the peak prefetched data misses during the last measurement period. Misses/sec

The detailed diagnosis shows when and how many prefetched data requests had been unsuccessful.