CosmosDbClientMetrics.HistogramBuckets.RequestLatencyBuckets Field
Definition
Important
Some information relates to prerelease product that may be substantially modified before it’s released. Microsoft makes no warranties, express or implied, with respect to the information provided here.
ExplicitBucketBoundaries for "db.client.operation.duration" Metrics.
public static readonly double[] RequestLatencyBuckets;
staticval mutable RequestLatencyBuckets : double[]
Public Shared ReadOnly RequestLatencyBuckets As Double()
Field Value
Remarks
0.001, 0.005, 0.010 seconds: Higher Precision at Sub-Millisecond Levels, For high-performance workloads, especially when dealing with microservices or low-latency queries.
0.050, 0.100, 0.200 seconds: Granularity for Standard Web Applications, These values allow detailed tracking for latencies between 50ms and 200ms, which are common in web applications. Fine-grained buckets in this range help in identifying performance issues before they grow critical, while covering the typical response times expected in Cosmos DB.
0.500, 1.000 seconds: Wider Range for Intermediate Latencies, Operations that take longer, in the range of 500ms to 1 second, are still important for performance monitoring. By capturing these values, you maintain awareness of potential bottlenecks or slower requests that may need optimization.
2.000, 5.000 seconds: Capturing Outliers and Slow Queries, It’s important to track outliers that might go beyond 1 second. Having buckets for 2 and 5 seconds enables identification of rare, long-running operations that may require further investigation.