Pipeline metrics example - Amazon Personalize

Pipeline metrics example

As you apply the Amazon Personalize Search Ranking plugin to OpenSearch queries, you can monitor the plugin by getting metrics for your search pipelines. Pipeline metrics include statistics like the number of failed requests for the personalized_search_ranking response processor.

The following code shows an excerpt of the pipeline metrics that are returned from OpenSearch. It shows only the pipelines object that contains statistics for two different pipelines. For each pipeline, you can find Amazon Personalize Search Ranking plugin metrics in the personalized_search_ranking response processor list. For a complete example of all metrics, see Search pipeline metrics.

{ .... .... "pipelines": { "pipelineA": { "request": { "count": 0, "time_in_millis": 0, "current": 0, "failed": 0 }, "response": { "count": 6, "time_in_millis": 2246, "current": 0, "failed": 0 }, "request_processors": [], "response_processors": [ { personalized_search_ranking": { "type": "personalized_search_ranking", "stats": { "count": <number of requests>, "time_in_millis": <time>, "current": 0, "failed": <number of failed requests> } } } ] }, "pipelineB": { "request": { "count": 0, "time_in_millis": 0, "current": 0, "failed": 0 }, "response": { "count": 8, "time_in_millis": 2248, "current": 0, "failed": 0 }, "request_processors": [], "response_processors": [ { "personalized_search_ranking": { "type": "personalized_search_ranking", "stats": { "count": <number of requests>, "time_in_millis": <time>, "current": 0, "failed": <number of failed requests> } } } ] } } .... .... }