You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Add air-gapped install instructions for Elasticsearch and Machine Learning (#2488)
* Add air-gapped install instructions for Elasticsearch and Machine Learning
* Add ToC entry for ML section
---------
Co-authored-by: David Kilfoyle <[email protected]>
@@ -49,7 +50,12 @@ Refer to the section for each Elastic component for air-gapped installation conf
49
50
[[air-gapped-elasticsearch]]
50
51
==== 1.1. {es}
51
52
52
-
Air-gapped install of {es} is fairly straightforward, as this component does not have any default dependencies on other services. Detailed install and configuration instructions are available in the {ref}/install-elasticsearch.html[{es} install documentation].
53
+
Air-gapped install of {es} may require additional steps in order to access some of the features. General install and configuration guides are available in the {ref}/install-elasticsearch.html[{es} install documentation].
54
+
55
+
Specifically:
56
+
57
+
* To be able to use the GeoIP processor, refer to {ref}/geoip-processor.html#manually-update-geoip-databases[the GeoIP processor documentation] for instructions on downloading and deploying the required databases.
58
+
* Refer to <<air-gapped-machine-learning,{ml-cap}>> for instructions on deploying the Elastic Learned Sparse EncodeR (ELSER) natural language processing (NLP) model and other trained {ml} models.
53
59
54
60
[discrete]
55
61
[[air-gapped-kibana]]
@@ -145,6 +151,15 @@ NOTE: When setting up own web server, such as NGINX, to function as the {artifac
145
151
146
152
Air-gapped setup of this component is, essentially, identical to the setup of the <<air-gapped-elastic-artifact-registry,{artifact-registry}>> except that different artifacts are served. To learn more, refer to {security-guide}/offline-endpoint.html[Configure offline endpoints and air-gapped environments] in the Elastic Security guide.
147
153
154
+
[discrete]
155
+
[[air-gapped-machine-learning]]
156
+
==== 1.13 {ml-cap}
157
+
158
+
Some {ml} features, like natural language processing (NLP), require you to deploy trained models. To learn about deploying {ml} models in an air-gapped environment, refer to:
159
+
160
+
* {ml-docs}/ml-nlp-elser.html#air-gapped-install[Deploy ELSER in an air-gapped environment].
161
+
* {eland-docs}/machine-learning.html#ml-nlp-pytorch-air-gapped[Install trained models in an air-gapped environment with Eland].
0 commit comments