k-NN plugin API

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

Stats

Introduced 1.0

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:

  1. GET /_plugins/_knn/nodeId1,nodeId2/stats/statName1,statName2
StatisticDescription
circuit_breaker_triggeredIndicates whether the circuit breaker is triggered. This statistic is only relevant to approximate k-NN search.
total_load_timeThe time in nanoseconds that k-NN has taken to load native library indices into the cache. This statistic is only relevant to approximate k-NN search.
eviction_countThe number of native library indices that have been evicted from the cache due to memory constraints or idle time. This statistic is only relevant to approximate k-NN search.
Note: Explicit evictions that occur because of index deletion aren’t counted.
hit_countThe number of cache hits. A cache hit occurs when a user queries a native library index that’s already loaded into memory. This statistic is only relevant to approximate k-NN search.
miss_countThe number of cache misses. A cache miss occurs when a user queries a native library index that isn’t loaded into memory yet. This statistic is only relevant to approximate k-NN search.
graph_memory_usageThe amount of native memory native library indices are using on the node in kilobytes.
graph_memory_usage_percentageThe amount of native memory native library indices are using on the node as a percentage of the maximum cache capacity.
graph_index_requestsThe number of requests to add the knn_vector field of a document into a native library index.
graph_index_errorsThe number of requests to add the knn_vector field of a document into a native library index that have produced an error.
graph_query_requestsThe number of native library index queries that have been made.
graph_query_errorsThe number of native library index queries that have produced an error.
knn_query_requestsThe number of k-NN query requests received.
cache_capacity_reachedWhether knn.memory.circuit_breaker.limit has been reached. This statistic is only relevant to approximate k-NN search.
load_success_countThe number of times k-NN successfully loaded a native library index into the cache. This statistic is only relevant to approximate k-NN search.
load_exception_countThe number of times an exception occurred when trying to load a native library index into the cache. This statistic is only relevant to approximate k-NN search.
indices_in_cacheFor each OpenSearch index with a knn_vector field and approximate k-NN turned on, this statistic provides the number of native library indices that OpenSearch index has and the total graph_memory_usage that the OpenSearch index is using, in kilobytes.
script_compilationsThe number of times the k-NN script has been compiled. This value should usually be 1 or 0, but if the cache containing the compiled scripts is filled, the k-NN script might be recompiled. This statistic is only relevant to k-NN score script search.
script_compilation_errorsThe number of errors during script compilation. This statistic is only relevant to k-NN score script search.
script_query_requestsThe total number of script queries. This statistic is only relevant to k-NN score script search.
script_query_errorsThe number of errors during script queries. This statistic is only relevant to k-NN score script search.
nmslib_initializedBoolean value indicating whether the nmslib JNI library has been loaded and initialized on the node.
faiss_initializedBoolean value indicating whether the faiss JNI library has been loaded and initialized on the node.
model_index_statusStatus of model system index. Valid values are “red”, “yellow”, “green”. If the index does not exist, this will be null.
indexing_from_model_degradedBoolean value indicating if indexing from a model is degraded. This will happen if there is not enough JVM memory to cache the models.
training_requestsThe number of training requests made to the node.
training_errorsThe number of training errors that have occurred on the node.
training_memory_usageThe amount of native memory training is using on the node in kilobytes.
training_memory_usage_percentageThe amount of native memory training is using on the node as a percentage of the maximum cache capacity.

Note: Some stats contain graph in the name. In these cases, graph is synonymous with native library index. The term graph is a legacy detail, coming from when the plugin only supported the HNSW algorithm, which consists of hierarchical graphs.

Usage

  1. GET /_plugins/_knn/stats?pretty
  2. {
  3. "_nodes" : {
  4. "total" : 1,
  5. "successful" : 1,
  6. "failed" : 0
  7. },
  8. "cluster_name" : "my-cluster",
  9. "circuit_breaker_triggered" : false,
  10. "model_index_status" : "YELLOW",
  11. "nodes" : {
  12. "JdfxIkOS1-43UxqNz98nw" : {
  13. "graph_memory_usage_percentage" : 3.68,
  14. "graph_query_requests" : 1420920,
  15. "graph_memory_usage" : 2,
  16. "cache_capacity_reached" : false,
  17. "load_success_count" : 179,
  18. "training_memory_usage" : 0,
  19. "indices_in_cache" : {
  20. "myindex" : {
  21. "graph_memory_usage" : 2,
  22. "graph_memory_usage_percentage" : 3.68,
  23. "graph_count" : 2
  24. }
  25. },
  26. "script_query_errors" : 0,
  27. "hit_count" : 1420775,
  28. "knn_query_requests" : 147092,
  29. "total_load_time" : 2436679306,
  30. "miss_count" : 179,
  31. "training_memory_usage_percentage" : 0.0,
  32. "graph_index_requests" : 656,
  33. "faiss_initialized" : true,
  34. "load_exception_count" : 0,
  35. "training_errors" : 0,
  36. "eviction_count" : 0,
  37. "nmslib_initialized" : false,
  38. "script_compilations" : 0,
  39. "script_query_requests" : 0,
  40. "graph_query_errors" : 0,
  41. "indexing_from_model_degraded" : false,
  42. "graph_index_errors" : 0,
  43. "training_requests" : 17,
  44. "script_compilation_errors" : 0
  45. }
  46. }
  47. }
  1. GET /_plugins/_knn/HYMrXXsBSamUkcAjhjeN0w/stats/circuit_breaker_triggered,graph_memory_usage?pretty
  2. {
  3. "_nodes" : {
  4. "total" : 1,
  5. "successful" : 1,
  6. "failed" : 0
  7. },
  8. "cluster_name" : "my-cluster",
  9. "circuit_breaker_triggered" : false,
  10. "nodes" : {
  11. "HYMrXXsBSamUkcAjhjeN0w" : {
  12. "graph_memory_usage" : 1
  13. }
  14. }
  15. }

Warmup operation

Introduced 1.0

The native library indices used to perform approximate k-Nearest Neighbor (k-NN) search are stored as special files with other Apache Lucene segment files. In order for you to perform a search on these indices using the k-NN plugin, the plugin needs to load these files into native memory.

If the plugin hasn’t loaded the files into native memory, it loads them when it receives a search request. The loading time can cause high latency during initial queries. To avoid this situation, 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 begin. 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 whatever indices you’re interested in searching. This operation loads all the native library files for all of the shards (primaries and replicas) of all the indices specified in the request into native memory.

After the process finishes, you can start searching against the indices with no 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 impact. It only loads files that aren’t currently in memory.

Usage

This request performs a warmup on three indices:

  1. GET /_plugins/_knn/warmup/index1,index2,index3?pretty
  2. {
  3. "_shards" : {
  4. "total" : 6,
  5. "successful" : 6,
  6. "failed" : 0
  7. }
  8. }

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 doesn’t return results until the warmup operation finishes or the request times out. If the request times out, the operation still continues on the cluster. To monitor the warmup operation, use the OpenSearch _tasks API:

  1. 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:

  • Don’t run merge operations on indices that you want to warm up. During merge, 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 indices A and B into native memory, but segment C is created from segments A and B being merged. The native library indices for 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 is still present.

  • Confirm that all native library indices 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.

  • Don’t 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 indices until they’re searchable. This means that you would have to run the warmup operation again after indexing finishes.

Get Model

Introduced 1.2

Used to retrieve 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 then be used to initialize native library index files during indexing. The model is serialized in the k-NN model system index.

  1. GET /_plugins/_knn/models/{model_id}
Response FieldDescription
model_idThe id of the fetched model.
model_blobThe base64 encoded string of the serialized model.
stateCurrent state of the model. Either “created”, “failed”, “training”.
timestampTime when the model was created.
descriptionUser provided description of the model.
errorError message explaining why the model is in the failed state.
space_typeSpace type this model is trained for.
dimensionDimension this model is for.
engineNative library used to create model. Either “faiss” or “nmslib”.

Usage

  1. GET /_plugins/_knn/models/test-model?pretty
  2. {
  3. "model_id" : "test-model",
  4. "model_blob" : "SXdGbIAAAAAAAAAAAA...",
  5. "state" : "created",
  6. "timestamp" : "2021-11-15T18:45:07.505369036Z",
  7. "description" : "Default",
  8. "error" : "",
  9. "space_type" : "l2",
  10. "dimension" : 128,
  11. "engine" : "faiss"
  12. }
  1. GET /_plugins/_knn/models/test-model?pretty&filter_path=model_id,state
  2. {
  3. "model_id" : "test-model",
  4. "state" : "created"
  5. }

Search Model

Introduced 1.2

Use an OpenSearch query to search for models in the index.

Usage

  1. GET/POST /_plugins/_knn/models/_search?pretty&_source_excludes=model_blob
  2. {
  3. "query": {
  4. ...
  5. }
  6. }
  7. {
  8. "took" : 0,
  9. "timed_out" : false,
  10. "_shards" : {
  11. "total" : 1,
  12. "successful" : 1,
  13. "skipped" : 0,
  14. "failed" : 0
  15. },
  16. "hits" : {
  17. "total" : {
  18. "value" : 1,
  19. "relation" : "eq"
  20. },
  21. "max_score" : 1.0,
  22. "hits" : [
  23. {
  24. "_index" : ".opensearch-knn-models",
  25. "_id" : "test-model",
  26. "_score" : 1.0,
  27. "_source" : {
  28. "engine" : "faiss",
  29. "space_type" : "l2",
  30. "description" : "Default",
  31. "model_id" : "test-model",
  32. "state" : "created",
  33. "error" : "",
  34. "dimension" : 128,
  35. "timestamp" : "2021-11-15T18:45:07.505369036Z"
  36. }
  37. }
  38. ]
  39. }
  40. }

Delete Model

Introduced 1.2

Used to delete a particular model in the cluster.

Usage

  1. DELETE /_plugins/_knn/models/{model_id}
  2. {
  3. "model_id": {model_id},
  4. "acknowledged": true
  5. }

Train Model

Introduced 1.2

Create and train a model that can be used for initializing k-NN native library indices during indexing. This API will pull training data from a knn_vector field in a training index and then create and train a model and then serialize it to the model system index. Training data must match the dimension passed into the body of the request. This request will return when training begins. To monitor the state of the model, use the Get model API.

Query ParameterDescription
model_id(Optional) The id of the fetched model. If not specified, a random id will be generated.
node_id(Optional) Preferred node to execute training. If set, this node will be used to perform training if it is deemed to be capable.
Request ParameterDescription
training_indexIndex from where training data from.
training_fieldknn_vector field from training_index to grab training data from. Dimension of this field must match dimension passed in to this request.
dimensionDimension this model is for.
max_training_vector_count(Optional) Maximum number of vectors from the training index to use for training. Defaults to all of the vectors in the index.
search_size(Optional) Training data is pulled from the training index with scroll queries. Defines the number of results to return per scroll query. Defaults to 10,000.
description(Optional) User provided description of the model.
methodConfiguration of ANN method used for search. For more information on possible methods, refer to the method documentation. Method must require training to be valid.

Usage

  1. POST /_plugins/_knn/models/{model_id}/_train?preference={node_id}
  2. {
  3. "training_index": "train-index-name",
  4. "training_field": "train-field-name",
  5. "dimension": 16,
  6. "max_training_vector_count": 1200,
  7. "search_size": 100,
  8. "description": "My model",
  9. "method": {
  10. "name":"ivf",
  11. "engine":"faiss",
  12. "space_type": "l2",
  13. "parameters":{
  14. "nlist":128,
  15. "encoder":{
  16. "name":"pq",
  17. "parameters":{
  18. "code_size":8
  19. }
  20. }
  21. }
  22. }
  23. }
  24. {
  25. "model_id": "model_x"
  26. }
  1. POST /_plugins/_knn/models/_train?preference={node_id}
  2. {
  3. "training_index": "train-index-name",
  4. "training_field": "train-field-name",
  5. "dimension": 16,
  6. "max_training_vector_count": 1200,
  7. "search_size": 100,
  8. "description": "My model",
  9. "method": {
  10. "name":"ivf",
  11. "engine":"faiss",
  12. "space_type": "l2",
  13. "parameters":{
  14. "nlist":128,
  15. "encoder":{
  16. "name":"pq",
  17. "parameters":{
  18. "code_size":8
  19. }
  20. }
  21. }
  22. }
  23. }
  24. {
  25. "model_id": "dcdwscddscsad"
  26. }