honeymoose/OpenSearch
1
0
Fork 0
mirror of https://github.com/honeymoose/OpenSearch.git synced 2025-02-08 22:14:59 +00:00
Code Issues Packages Projects Releases Wiki Activity

[DOCS] Document static machine learning settings (#61382)

Browse Source
This commit is contained in:
Lisa Cawley 2020-08-24 07:29:25 -07:00 committed by lcawl
parent af2e2782eb
commit 52b12a07c4
2 changed files with 91 additions and 93 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

179
docs/reference/settings/ml-settings.asciidoc
View File

@ -10,15 +10,9 @@
// tag::ml-settings-description-tag[] // tag::ml-settings-description-tag[]
You do not need to configure any settings to use {ml}. It is enabled by default. You do not need to configure any settings to use {ml}. It is enabled by default.
IMPORTANT: {ml-cap} uses SSE4.2 instructions, so will only work on machines whose IMPORTANT: {ml-cap} uses SSE4.2 instructions, so it works only on machines whose
CPUs {wikipedia}/SSE4#Supporting_CPUs[support] SSE4.2. If you CPUs {wikipedia}/SSE4#Supporting_CPUs[support] SSE4.2. If you run {es} on older
run {es} on older hardware you must disable {ml} (by setting `xpack.ml.enabled` hardware, you must disable {ml} (by setting `xpack.ml.enabled` to `false`).
to `false`).
All of these settings can be added to the `elasticsearch.yml` configuration
file. The dynamic settings can also be updated across a cluster with the
<<cluster-update-settings,cluster update settings API>>. Dynamic settings take
precedence over settings in the `elasticsearch.yml` file.
// end::ml-settings-description-tag[] // end::ml-settings-description-tag[]
@ -27,8 +21,9 @@ precedence over settings in the `elasticsearch.yml` file.
==== General machine learning settings ==== General machine learning settings
`node.roles: [ ml ]`:: `node.roles: [ ml ]`::
Set `node.roles` to contain `ml` to identify the node as a _{ml} node_ that is (<<static-cluster-setting,Static>>) Set `node.roles` to contain `ml` to identify
capable of running jobs. Every node is a {ml} node by default.+ the node as a _{ml} node_ that is capable of running jobs. Every node is a {ml}
node by default.
+ +
If you use the `node.roles` setting, then all required roles must be explicitly If you use the `node.roles` setting, then all required roles must be explicitly
set. Consult <<modules-node>> to learn more. set. Consult <<modules-node>> to learn more.
@ -38,7 +33,8 @@ the `ml` role.
`xpack.ml.enabled`:: `xpack.ml.enabled`::
Set to `true` (default) to enable {ml} APIs on the node. (<<static-cluster-setting,Static>>) Set to `true` (default) to enable {ml} APIs
on the node.
+ +
If set to `false`, the {ml} APIs are disabled on the node. Therefore the node If set to `false`, the {ml} APIs are disabled on the node. Therefore the node
cannot open jobs, start {dfeeds}, or receive transport (internal) communication cannot open jobs, start {dfeeds}, or receive transport (internal) communication
@ -54,58 +50,62 @@ want to use {ml-features} in clients or {kib}, it must also be enabled on all
coordinating nodes. coordinating nodes.
`xpack.ml.inference_model.cache_size`:: `xpack.ml.inference_model.cache_size`::
The maximum inference cache size allowed. The inference cache exists in the JVM (<<static-cluster-setting,Static>>) The maximum inference cache size allowed.
heap on each ingest node. The cache affords faster processing times for the The inference cache exists in the JVM heap on each ingest node. The cache
`inference` processor. The value can be a static byte sized value (i.e. "2gb") affords faster processing times for the `inference` processor. The value can be
or a percentage of total allocated heap. The default is "40%". a static byte sized value (i.e. "2gb") or a percentage of total allocated heap.
See also <<model-inference-circuit-breaker>>. The default is "40%". See also <<model-inference-circuit-breaker>>.
[[xpack-interference-model-ttl]] [[xpack-interference-model-ttl]]
// tag::interference-model-ttl-tag[] // tag::interference-model-ttl-tag[]
`xpack.ml.inference_model.time_to_live` {ess-icon}:: `xpack.ml.inference_model.time_to_live` {ess-icon}::
The time to live (TTL) for models in the inference model cache. The TTL is (<<static-cluster-setting,Static>>) The time to live (TTL) for models in the
calculated from last access. The `inference` processor attempts to load the inference model cache. The TTL is calculated from last access. The `inference`
model from cache. If the `inference` processor does not receive any documents processor attempts to load the model from cache. If the `inference` processor
for the duration of the TTL, the referenced model is flagged for eviction from does not receive any documents for the duration of the TTL, the referenced model
the cache. If a document is processed later, the model is again loaded into the is flagged for eviction from the cache. If a document is processed later, the
cache. Defaults to `5m`. model is again loaded into the cache. Defaults to `5m`.
// end::interference-model-ttl-tag[] // end::interference-model-ttl-tag[]
`xpack.ml.max_inference_processors` (<<cluster-update-settings,Dynamic>>):: `xpack.ml.max_inference_processors`::
The total number of `inference` type processors allowed across all ingest (<<cluster-update-settings,Dynamic>>) The total number of `inference` type
pipelines. Once the limit is reached, adding an `inference` processor to processors allowed across all ingest pipelines. Once the limit is reached,
a pipeline is disallowed. Defaults to `50`. adding an `inference` processor to a pipeline is disallowed. Defaults to `50`.
`xpack.ml.max_machine_memory_percent` (<<cluster-update-settings,Dynamic>>):: `xpack.ml.max_machine_memory_percent`::
The maximum percentage of the machine's memory that {ml} may use for running (<<cluster-update-settings,Dynamic>>) The maximum percentage of the machine's
analytics processes. (These processes are separate to the {es} JVM.) Defaults to memory that {ml} may use for running analytics processes. (These processes are
`30` percent. The limit is based on the total memory of the machine, not current separate to the {es} JVM.) Defaults to `30` percent. The limit is based on the
free memory. Jobs will not be allocated to a node if doing so would cause the total memory of the machine, not current free memory. Jobs are not allocated to
estimated memory use of {ml} jobs to exceed the limit. a node if doing so would cause the estimated memory use of {ml} jobs to exceed
the limit.
`xpack.ml.max_model_memory_limit` (<<cluster-update-settings,Dynamic>>):: `xpack.ml.max_model_memory_limit`::
The maximum `model_memory_limit` property value that can be set for any job on (<<cluster-update-settings,Dynamic>>) The maximum `model_memory_limit` property
this node. If you try to create a job with a `model_memory_limit` property value value that can be set for any job on this node. If you try to create a job with
that is greater than this setting value, an error occurs. Existing jobs are not a `model_memory_limit` property value that is greater than this setting value,
affected when you update this setting. For more information about the an error occurs. Existing jobs are not affected when you update this setting.
`model_memory_limit` property, see <<put-analysislimits>>. For more information about the `model_memory_limit` property, see
<<put-analysislimits>>.
[[xpack.ml.max_open_jobs]] [[xpack.ml.max_open_jobs]]
`xpack.ml.max_open_jobs` (<<cluster-update-settings,Dynamic>>):: `xpack.ml.max_open_jobs`::
The maximum number of jobs that can run simultaneously on a node. Defaults to (<<cluster-update-settings,Dynamic>>) The maximum number of jobs that can run
`20`. In this context, jobs include both {anomaly-jobs} and {dfanalytics-jobs}. simultaneously on a node. Defaults to `20`. In this context, jobs include both
The maximum number of jobs is also constrained by memory usage. Thus if the {anomaly-jobs} and {dfanalytics-jobs}. The maximum number of jobs is also
estimated memory usage of the jobs would be higher than allowed, fewer jobs will constrained by memory usage. Thus if the estimated memory usage of the jobs
run on a node. Prior to version 7.1, this setting was a per-node non-dynamic would be higher than allowed, fewer jobs will run on a node. Prior to version
setting. It became a cluster-wide dynamic setting in version 7.1. As a result, 7.1, this setting was a per-node non-dynamic setting. It became a cluster-wide
changes to its value after node startup are used only after every node in the dynamic setting in version 7.1. As a result, changes to its value after node
cluster is running version 7.1 or higher. The maximum permitted value is `512`. startup are used only after every node in the cluster is running version 7.1 or
higher. The maximum permitted value is `512`.
`xpack.ml.node_concurrent_job_allocations` (<<cluster-update-settings,Dynamic>>):: `xpack.ml.node_concurrent_job_allocations`::
The maximum number of jobs that can concurrently be in the `opening` state on (<<cluster-update-settings,Dynamic>>) The maximum number of jobs that can
each node. Typically, jobs spend a small amount of time in this state before concurrently be in the `opening` state on each node. Typically, jobs spend a
they move to `open` state. Jobs that must restore large models when they are small amount of time in this state before they move to `open` state. Jobs that
opening spend more time in the `opening` state. Defaults to `2`. must restore large models when they are opening spend more time in the `opening`
state. Defaults to `2`.
[discrete] [discrete]
[[advanced-ml-settings]] [[advanced-ml-settings]]
@ -114,52 +114,55 @@ opening spend more time in the `opening` state. Defaults to `2`.
These settings are for advanced use cases; the default values are generally These settings are for advanced use cases; the default values are generally
sufficient: sufficient:
`xpack.ml.enable_config_migration` (<<cluster-update-settings,Dynamic>>):: `xpack.ml.enable_config_migration`::
Reserved. (<<cluster-update-settings,Dynamic>>) Reserved.
`xpack.ml.max_anomaly_records` (<<cluster-update-settings,Dynamic>>):: `xpack.ml.max_anomaly_records`::
The maximum number of records that are output per bucket. The default value is (<<cluster-update-settings,Dynamic>>) The maximum number of records that are
`500`. output per bucket. The default value is `500`.
`xpack.ml.max_lazy_ml_nodes` (<<cluster-update-settings,Dynamic>>):: `xpack.ml.max_lazy_ml_nodes`::
The number of lazily spun up Machine Learning nodes. Useful in situations (<<cluster-update-settings,Dynamic>>) The number of lazily spun up {ml} nodes.
where ML nodes are not desired until the first Machine Learning Job Useful in situations where {ml} nodes are not desired until the first {ml} job
is opened. It defaults to `0` and has a maximum acceptable value of `3`. opens. It defaults to `0` and has a maximum acceptable value of `3`. If the
If the current number of ML nodes is `>=` than this setting, then it is current number of {ml} nodes is greater than or equal to this setting, it is
assumed that there are no more lazy nodes available as the desired number assumed that there are no more lazy nodes available as the desired number
of nodes have already been provisioned. When a job is opened with this of nodes have already been provisioned. If a job is opened and this setting has
setting set at `>0` and there are no nodes that can accept the job, then a value greater than zero and there are no nodes that can accept the job, the
the job will stay in the `OPENING` state until a new ML node is added to the job stays in the `OPENING` state until a new {ml} node is added to the cluster
cluster and the job is assigned to run on that node. and the job is assigned to run on that node.
+ +
IMPORTANT: This setting assumes some external process is capable of adding ML nodes IMPORTANT: This setting assumes some external process is capable of adding {ml}
to the cluster. This setting is only useful when used in conjunction with nodes to the cluster. This setting is only useful when used in conjunction with
such an external process. such an external process.
`xpack.ml.process_connect_timeout` (<<cluster-update-settings,Dynamic>>):: `xpack.ml.process_connect_timeout`::
The connection timeout for {ml} processes that run separately from the {es} JVM. (<<cluster-update-settings,Dynamic>>) The connection timeout for {ml} processes
Defaults to `10s`. Some {ml} processing is done by processes that run separately that run separately from the {es} JVM. Defaults to `10s`. Some {ml} processing
to the {es} JVM. When such processes are started they must connect to the {es} is done by processes that run separately to the {es} JVM. When such processes
JVM. If such a process does not connect within the time period specified by this are started they must connect to the {es} JVM. If such a process does not
setting then the process is assumed to have failed. Defaults to `10s`. The minimum connect within the time period specified by this setting then the process is
value for this setting is `5s`. assumed to have failed. Defaults to `10s`. The minimum value for this setting is
`5s`.
[discrete] [discrete]
[[model-inference-circuit-breaker]] [[model-inference-circuit-breaker]]
==== {ml-cap} circuit breaker settings ==== {ml-cap} circuit breaker settings
`breaker.model_inference.limit` (<<cluster-update-settings,Dynamic>>):: `breaker.model_inference.limit`::
Limit for the model inference breaker, which defaults to 50% of the JVM heap. (<<cluster-update-settings,Dynamic>>) Limit for the model inference breaker,
If the parent circuit breaker is less than 50% of the JVM heap, it is bound which defaults to 50% of the JVM heap. If the parent circuit breaker is less
to that limit instead. See <<circuit-breaker>>. than 50% of the JVM heap, it is bound to that limit instead. See
<<circuit-breaker>>.
`breaker.model_inference.overhead` (<<cluster-update-settings,Dynamic>>):: `breaker.model_inference.overhead`::
A constant that all accounting estimations are multiplied by to determine (<<cluster-update-settings,Dynamic>>) A constant that all accounting estimations
a final estimation. Defaults to 1. See <<circuit-breaker>>. are multiplied by to determine a final estimation. Defaults to 1. See
<<circuit-breaker>>.
`breaker.model_inference.type`:: `breaker.model_inference.type`::
The underlying type of the circuit breaker. There are two valid options: `noop` (<<static-cluster-setting,Static>>) The underlying type of the circuit breaker.
and `memory`. `noop` means the circuit breaker does nothing to prevent too much There are two valid options: `noop` and `memory`. `noop` means the circuit
memory usage. `memory` means the circuit breaker tracks the memory used by breaker does nothing to prevent too much memory usage. `memory` means the
inference models and can potentially break and prevent OutOfMemory errors. The circuit breaker tracks the memory used by inference models and can potentially
default is `memory`. break and prevent `OutOfMemory` errors. The default is `memory`.

5
docs/reference/settings/transform-settings.asciidoc
View File

@ -10,11 +10,6 @@
You do not need to configure any settings to use {transforms}. It is enabled by You do not need to configure any settings to use {transforms}. It is enabled by
default. default.
All of these settings can be added to the `elasticsearch.yml` configuration file.
The dynamic settings can also be updated across a cluster with the
<<cluster-update-settings,cluster update settings API>>. Dynamic settings take
precedence over settings in the `elasticsearch.yml` file.
[discrete] [discrete]
[[general-transform-settings]] [[general-transform-settings]]
==== General {transforms} settings ==== General {transforms} settings