k-NN API

Preview

The k-NN plugin adds several APIs for managing, monitoring, and optimizing your k-NN workload.

Stats

The k-NN stats API provides information about the current status of the k-NN plugin. The plugin keeps track of both cluster-level and node-level statistics. Cluster-level statistics have a single value for the entire cluster. Node-level statistics have a single value for each node in the cluster. You can filter the query by nodeId and statName, as shown in the following example:

GET /_plugins/_knn/nodeId1,nodeId2/stats/statName1,statName2

Some statistics contain graph in the name. In these cases, graph is synonymous with native library index. The term graph is reflective of when the plugin only supported the HNSW algorithm, which consists of hierarchical graphs.

Usage

The following code examples show how to retrieve statistics related to the k-NN plugin. The first example fetches comprehensive statistics for the k-NN plugin across all nodes in the cluster, while the second example retrieves specific metrics (circuit breaker status and graph memory usage) for a single node.

GET /_plugins/_knn/stats?pretty
{
    "_nodes" : {
        "total" : 1,
        "successful" : 1,
        "failed" : 0
    },
    "cluster_name" : "my-cluster",
    "circuit_breaker_triggered" : false,
    "model_index_status" : "YELLOW",
    "nodes" : {
      "JdfxIkOS1-43UxqNz98nw" : {
        "graph_memory_usage_percentage" : 3.68,
        "graph_query_requests" : 1420920,
        "graph_memory_usage" : 2,
        "cache_capacity_reached" : false,
        "load_success_count" : 179,
        "training_memory_usage" : 0,
        "indices_in_cache" : {
            "myindex" : {
                "graph_memory_usage" : 2,
                "graph_memory_usage_percentage" : 3.68,
                "graph_count" : 2
            }
        },
        "script_query_errors" : 0,
        "hit_count" : 1420775,
        "knn_query_requests" : 147092,
        "total_load_time" : 2436679306,
        "miss_count" : 179,
        "training_memory_usage_percentage" : 0.0,
        "graph_index_requests" : 656,
        "faiss_initialized" : true,
        "load_exception_count" : 0,
        "training_errors" : 0,
        "eviction_count" : 0,
        "nmslib_initialized" : false,
        "script_compilations" : 0,
        "script_query_requests" : 0,
        "graph_query_errors" : 0,
        "indexing_from_model_degraded" : false,
        "graph_index_errors" : 0,
        "training_requests" : 17,
        "script_compilation_errors" : 0
    }
  }
}

GET /_plugins/_knn/HYMrXXsBSamUkcAjhjeN0w/stats/circuit_breaker_triggered,graph_memory_usage?pretty
{
    "_nodes" : {
        "total" : 1,
        "successful" : 1,
        "failed" : 0
    },
    "cluster_name" : "my-cluster",
    "circuit_breaker_triggered" : false,
    "nodes" : {
        "HYMrXXsBSamUkcAjhjeN0w" : {
            "graph_memory_usage" : 1
        }
    }
}

Warmup operation

The native library indexes used to perform approximate k-NN search are stored as special files with other Apache Lucene segment files. To perform a search on these indexes using the k-NN plugin, the plugin needs to load these files into native memory.

If the plugin has not loaded the files into native memory, then it loads them when it receives a search request. The loading time can cause high latency during initial queries. To avoid this, users often run random queries during a warmup period. After this warmup period, the files are loaded into native memory, and their production workloads can launch. This loading process is indirect and requires extra effort.

As an alternative, you can avoid this latency issue by running the k-NN plugin warmup API operation on the indexes you want to search. This operation loads all the native library files for all the shards (primaries and replicas) of all the indexes specified in the request into native memory.

After the process is finished, you can search against the indexes without initial latency penalties. The warmup API operation is idempotent, so if a segment’s native library files are already loaded into memory, this operation has no effect. It only loads files not currently stored in memory.

Usage

This request performs a warmup on three indexes:

GET /_plugins/_knn/warmup/index1,index2,index3?pretty
{
  "_shards" : {
    "total" : 6,
    "successful" : 6,
    "failed" : 0
  }
}

total indicates how many shards the k-NN plugin attempted to warm up. The response also includes the number of shards the plugin succeeded and failed to warm up.

The call does not return results until the warmup operation finishes or the request times out. If the request times out, then the operation continues on the cluster. To monitor the warmup operation, use the Lucenia _tasks API:

GET /_tasks

After the operation has finished, use the k-NN _stats API operation to see what the k-NN plugin loaded into the graph.

Best practices

For the warmup operation to function properly, follow these best practices:

• Do not run merge operations on indexes that you want to warm up. During a merge operation, the k-NN plugin creates new segments, and old segments are sometimes deleted. For example, you could encounter a situation in which the warmup API operation loads native library indexes A and B into native memory but segment C is created from segments A and B being merged. Native library indexes A and B would no longer be in memory, and native library index C would also not be in memory. In this case, the initial penalty for loading native library index C still exists.

• Confirm that all native library indexes you want to warm up can fit into native memory. For more information about the native memory limit, see the knn.memory.circuit_breaker.limit statistic. High graph memory usage causes cache thrashing, which can lead to operations constantly failing and attempting to run again.

• Do not index any documents that you want to load into the cache. Writing new information to segments prevents the warmup API operation from loading the native library indexes until they are searchable. This means that you would have to run the warmup operation again after indexing.

k-NN clear cache

Introduced 0.1.0

During approximate k-NN search or warmup operations, the native library indexes (nmslib and faiss engines) are loaded into native memory. Currently, you can evict an index from cache or native memory by either deleting the index or setting the k-NN cluster settings knn.cache.item.expiry.enabled and knn.cache.item.expiry.minutes, which removes the index from the cache if it is idle for a given period of time. However, you cannot evict an index from the cache without deleting the index. To solve this problem, you can use the k-NN clear cache API operation, which clears a given set of indexes from the cache.

The k-NN clear cache API evicts all native library files for all shards (primaries and replicas) of all indexes specified in the request. Similarly to how the warmup operation behaves, the k-NN clear cache API is idempotent, meaning that if you try to clear the cache for an index that has already been evicted from the cache, it does not have any additional effect.

This API operation only works with indexes created using the nmslib and faiss engines. It has no effect on indexes created using the lucene engine.

Usage

The following request evicts the native library indexes of three indexes from the cache:

GET /_plugins/_knn/clear_cache/index1,index2,index3?pretty
{
  "_shards" : {
    "total" : 6,
    "successful" : 6,
    "failed" : 0
  }
}

The total parameter indicates the number of shards that the API attempted to clear from the cache. The response includes both the number of cleared shards and the number of shards that the plugin failed to clear.

The k-NN clear cache API can be used with index patterns to clear one or more indexes that match the given pattern from the cache, as shown in the following example:

GET /_plugins/_knn/clear_cache/index*?pretty
{
  "_shards" : {
    "total" : 6,
    "successful" : 6,
    "failed" : 0
  }
}

The API call does not return results until the operation finishes or the request times out. If the request times out, then the operation continues on the cluster. To monitor the request, use the _tasks API, as shown in the following example:

GET /_tasks

When the operation finishes, use the k-NN _stats API operation to see which indexes have been evicted from the cache.

Get a model

The GET model operation retrieves information about models present in the cluster. Some native library index configurations require a training step before indexing and querying can begin. The output of training is a model that can be used to initialize native library index files during indexing. The model is serialized in the k-NN model system index. See the following GET example:

GET /_plugins/_knn/models/{model_id}

Usage

The following examples show how to retrieve information about a specific model using the k-NN plugin API. The first example returns all the available information about the model, while the second example shows how to selectively retrieve fields.

GET /_plugins/_knn/models/test-model?pretty
{
  "model_id" : "test-model",
  "model_blob" : "SXdGbIAAAAAAAAAAAA...",
  "state" : "created",
  "timestamp" : "2021-11-15T18:45:07.505369036Z",
  "description" : "Default",
  "error" : "",
  "space_type" : "l2",
  "dimension" : 128,
  "engine" : "faiss" 
}

GET /_plugins/_knn/models/test-model?pretty&filter_path=model_id,state
{
  "model_id" : "test-model",
  "state" : "created"
}

Search for a model

You can use an Lucenia query to search for a model in the index. See the following usage example.

Usage

The following example shows how to search for k-NN models in an Lucenia cluster and how to retrieve the metadata for those models, excluding the potentially large model_blob field:

GET/POST /_plugins/_knn/models/_search?pretty&_source_excludes=model_blob
{
    "query": {
         ...
     }
}

{
    "took" : 0,
    "timed_out" : false,
    "_shards" : {
        "total" : 1,
        "successful" : 1,
        "skipped" : 0,
        "failed" : 0
    },
    "hits" : {
      "total" : {
          "value" : 1,
          "relation" : "eq"
      },
    "max_score" : 1.0,
    "hits" : [
      {
        "_index" : ".lucenia-knn-models",
        "_id" : "test-model",
        "_score" : 1.0,
        "_source" : {
          "engine" : "faiss",
          "space_type" : "l2",
          "description" : "Default",
          "model_id" : "test-model",
          "state" : "created",
          "error" : "",
          "dimension" : 128,
          "timestamp" : "2021-11-15T18:45:07.505369036Z"
        }
      }
    ]
  }
}

Delete a model

You can delete a model in the cluster by using the DELETE operation. See the following usage example.

Usage

The following example shows how to delete a k-NN model:

DELETE /_plugins/_knn/models/{model_id}
{
  "model_id": {model_id},
  "acknowledged": true
}

Train a model

You can create and train a model that can be used for initializing k-NN native library indexes during indexing. This API pulls training data from a knn_vector field in a training index, creates and trains a model, and then serializes it to the model system index. Training data must match the dimension passed in the request body. This request is returned when training begins. To monitor the model’s state, use the Get model API.

Usage

The following examples show how to initiate the training process for a k-NN model:

POST /_plugins/_knn/models/{model_id}/_train?preference={node_id}
{
    "training_index": "train-index-name",
    "training_field": "train-field-name",
    "dimension": 16,
    "max_training_vector_count": 1200,
    "search_size": 100,
    "description": "My model",
    "method": {
        "name":"ivf",
        "engine":"faiss",
        "space_type": "l2",
        "parameters":{
            "nlist":128,
            "encoder":{
                "name":"pq",
                "parameters":{
                    "code_size":8
                }
            }
        }
    }
}

{
    "model_id": "model_x"
}

POST /_plugins/_knn/models/_train?preference={node_id}
{
    "training_index": "train-index-name",
    "training_field": "train-field-name",
    "dimension": 16,
    "max_training_vector_count": 1200,
    "search_size": 100,
    "description": "My model",
    "method": {
        "name":"ivf",
        "engine":"faiss",
        "space_type": "l2",
        "parameters":{
            "nlist":128,
            "encoder":{
                "name":"pq",
                "parameters":{
                    "code_size":8
                }
            }
        }
    }
}

{
    "model_id": "dcdwscddscsad"
}