eG Monitoring
 

Measures reported by SybDataCacheTest

This test reports the statistics pertaining to the default data cache and for all the named cahes configured in a Sybase database server. The measures made by this test are as follows:

Measurement Description Measurement Unit Interpretation
Utilization The percentage of searches using this cache as a percentage of searches across all caches Percent Compare this value for each cache to determine if there are caches that are over or under-utilized. If you decide that a cache is not well utilized, you can
  • Change the cache bindings to balance utilization
  • Resize the cache to correspond more appropriately to its utilization
Cache_hit_ratio The percentage of times that a required page was found in the data cache Percent A low value indicates that the cache may be too small for the current application load.
Spinlock_contention The number of times an engine encountered spinlock contention on the cache, and had to wait, as a percentage of the total spinlock requests for that cache Percent If spinlock contention is more than 10%, consider using named caches or adding cache partitions.
Buffers_grabbed The average number of times that Adaptive Server fetched a buffer from the LRU end of the cache, replacing a database page Buffers/Sec The detailed diagnosis of this measure, if enabled, reports the buffers grabbed for each configured memory pool. If this value is high in some pools and low in other pools, you might want to move space from the less active pool to the more active pool, especially if it can improve the cache-hit ratio
Buffers_grabbed_dirty The average number of times that fetching a buffer found a dirty page at the LRU end of the cache and had to wait while the buffer was written to disk Buffers/Sec A nonzero value indicates that the wash area of the pool is too small for the throughput in the pool.The detailed diagnosis of this measure, if enabled, reports the dirty buffers grabbed for each configured memory pool.
Pct_large_io_denied The number of large I/O requests denied as a percentage of the total number of requests made Percent Large I/O requests are denied due to the following reasons:
  • If any page in a buffer already resides in another pool.
  • When there are no buffers available in the requested pool.
  • On the first extent of an allocation unit, since it contains the allocation page, which is always read into the 2K pool.
If a high percentage of large I/Os were denied, it indicates that the use of the large pools might not be as effective as it could be.
Large_io_effectiveness The total number of pages that were used after being brought into cache by large I/O Percent A low value indicates that only few of the pages brought into cache are being accessed by queries.