OSS accelerator monitoring provides four metrics—Bandwidth Usage, QPS Usage, Hit Rate, and Latency—to help you assess accelerator performance and identify opportunities to improve cache efficiency.
Prerequisites
Before you begin, ensure that you have:
An OSS accelerator — see Create an OSS accelerator
Monitoring items
The following table describes each monitoring item and what it measures.
| Monitoring item | What it measures |
|---|---|
| Bandwidth Usage | Average outbound traffic over a given period, indicating the overall data transfer rate |
| QPS Usage | Average number of requests per second over a given period, indicating how frequently the accelerator is accessed |
| Hit Rate | Percentage of data served from the accelerator cache out of total data downloaded in a given period. A higher value indicates more efficient cache utilization |
| Latency | Average time to serve cached data, calculated as total transfer time for cache hits divided by total number of hits in a given period |
View monitoring data
Go to the Buckets page and select the bucket to manage.
In the left-side navigation tree, choose Bucket Settings > OSS Accelerator. In the accelerator list, click View Details in the Actions column to go to the details page of accelerator, then click Accelerator Monitoring.
Specify a time range and click Search.
The chart displays five metrics. The following table describes each one.
| Metric | Description | Unit |
|---|---|---|
| Outbound Bandwidth from OSS Accelerator | Average outbound traffic rate served from the accelerator cache | Kbps |
| Outbound Bandwidth from Origin | Average outbound traffic rate of a standard OSS bucket | Kbps |
| Read Request | Number of read requests per second | Requests/s |
| Hit Rate | Hit rate for downloading accelerated cache data | % |
| Cache Access | Average time to serve data from the accelerator cache | ms |
Reading the bandwidth metrics together: Outbound Bandwidth from OSS Accelerator is traffic served from cache. Outbound Bandwidth from Origin is traffic that bypassed the cache and was fetched from the origin bucket—this traffic incurs origin egress charges. A high origin bandwidth relative to accelerator bandwidth indicates a low cache hit rate.
Improve the cache hit rate
When a request is not served from cache, the accelerator fetches the object from the origin bucket (a cache miss). Cache misses increase latency and origin egress costs. If Hit Rate is lower than expected, you can change the path of the accelerator to cache more data and expand the cache capacity. This prevents frequent data elimination and improves transfer performance.
| Condition | Action |
|---|---|
| The accelerator path covers only a subset of the objects your users access | Change the accelerator path to cover a broader set of frequently accessed objects, so more data is eligible for caching |
| Cached objects are being evicted under high-traffic conditions | Expand the cache capacity to retain more objects and reduce eviction frequency |