github-mirrors/elasticsearch
1
0
Fork 0
mirror of https://github.com/elastic/elasticsearch.git synced 2025-06-29 18:03:32 -04:00
Code Issues Projects Releases Packages Wiki Activity
elasticsearch/docs/reference
History
Benjamin Trent afa28d49b4
[ML] add new cache_size parameter to trained_model deployments API (#88450) 
With: https://github.com/elastic/ml-cpp/pull/2305 we now support caching pytorch inference responses per node per model.

By default, the cache will be the same size has the model on disk size. This is because our current best estimate for memory used (for deploying) is 2*model_size + constant_overhead. 

This is due to the model having to be loaded in memory twice when serializing to the native process. 

But, once the model is in memory and accepting requests, its actual memory usage is reduced vs. what we have "reserved" for it within the node.

Consequently, having a cache layer that takes advantage of that unused (but reserved) memory is effectively free. When used in production, especially in search scenarios, caching inference results is critical for decreasing latency.
2022-07-18 09:19:01 -04:00
..
aggregations Corrected an incomplete sentence. (#86542) 2022-07-12 09:19:58 -04:00
analysis [DOCS] Add note for tokenizers that don't support keep types token filter (#87553) 2022-06-13 11:28:32 +02:00
autoscaling [DOCS] Document autoscaling processors (#88248) 2022-07-05 13:51:51 +02:00
cat Move the ingest attachment processor to the default distribution (#87989) 2022-06-28 02:10:36 -04:00
ccr [DOCS] Clarify when changes are replicated in CCR (#83863) 2022-02-11 16:07:38 -05:00
cluster Improve description for task api detailed param (#88493) 2022-07-14 09:22:28 +02:00
commands Fix some typos in plugins & reference docs (#84667) 2022-03-07 12:29:58 -05:00
data-management [DOC] auto migrate only for default template (#82043) 2022-05-10 11:35:19 -04:00
data-streams Added additional index.look_ahead_time validation (#87847) 2022-06-21 10:50:33 +02:00
docs Docs: Data streams only support create (#87263) 2022-06-08 13:41:42 -04:00
eql [DOCS] Fix ignore_unavailable parameter definition (#84071) 2022-02-17 08:24:06 -05:00
features/apis Make feature reset API response more informative (#71240) 2021-04-27 13:47:10 -04:00
fleet Fix some typos in plugins & reference docs (#84667) 2022-03-07 12:29:58 -05:00
graph [DOCS] Fix typos (#83895) 2022-02-15 12:42:17 -05:00
health Adding more master_is_stable details (#87977) 2022-06-28 09:21:21 -05:00
high-availability [DOCS] Overhaul snapshot and restore docs (#79081) 2021-11-15 12:45:07 -05:00
how-to [DOCS] Warn about impact of large readahead on search (#88007) 2022-06-27 13:00:44 +03:00
ilm Fix byte unit typo in ILM rollover doc (#87780) 2022-06-23 09:33:47 +02:00
images Add troubleshooting guide for corrupt repository (#88391) 2022-07-14 13:37:02 +01:00
index-modules [DOCS] Fix typos in docs (#88226) 2022-07-05 11:02:29 +02:00
indices [DOCS] Add TSDS docs, take two (#87703) 2022-06-16 12:44:10 -04:00
ingest [DOCS] Fixes a link that breaks the docs build. (#88111) 2022-06-28 10:22:23 +02:00
licensing [DOCS] Remove testenv annotations from doc snippet tests (#80023) 2021-11-05 18:38:50 -04:00
mapping Add 'mode' option to _source field mapper (#88211) 2022-07-18 12:50:10 +01:00
migration Forward-port 8.3.1 release notes 2022-07-01 11:19:14 +01:00
ml [ML] add new cache_size parameter to trained_model deployments API (#88450) 2022-07-18 09:19:01 -04:00
modules [DOCS] Adding discovery troubleshooting link in the master get help page (#87344) 2022-07-06 15:51:43 -04:00
monitoring [DOCS] Adding Getting Help section to troubleshooting docs (#87095) 2022-05-25 15:58:41 -04:00
query-dsl Undeprecate function_score query (#87807) 2022-06-17 11:04:26 -04:00
release-notes [DOCS] Added 8.3.2 Release Notes stub (#88332) 2022-07-07 08:41:22 -05:00
repositories-metering-api [DOCS] Remove testenv annotations from doc snippet tests (#80023) 2021-11-05 18:38:50 -04:00
rest-api Remove suggest flag from index stats docs (#85479) 2022-07-14 12:50:53 -04:00
rollup [DOCS] Remove testenv annotations from doc snippet tests (#80023) 2021-11-05 18:38:50 -04:00
scripting [Docs] Fix runtime grok script example (#87851) 2022-07-05 10:53:24 -04:00
search Remove Collector implementation from BucketCollector (#88444) 2022-07-18 08:18:13 +02:00
searchable-snapshots Add note that searchable snapshots indices cannot be snapshotted into source-only repositories (#86208) 2022-05-06 11:33:50 +02:00
settings Add setting for tcp_keepalive for oidc back-channel (#87868) 2022-07-07 11:41:14 +09:30
setup Add build_flavor back to info api rest response (#88336) 2022-07-08 09:54:29 +09:30
shutdown/apis [doc] Explicitly mention about node shutdown remove for cluster shrink (#86173) 2022-05-09 10:24:54 +02:00
slm/apis [DOCS] Remove soft limit for snapshot repositories (#80745) 2021-11-16 12:24:18 -05:00
snapshot-restore Clarify snapshot docs on archive indices (#88417) 2022-07-11 12:03:18 +02:00
sql [DOCS] Fix typos in docs (#88226) 2022-07-05 11:02:29 +02:00
tab-widgets Add troubleshooting guide for corrupt repository (#88391) 2022-07-14 13:37:02 +01:00
text-structure/apis [DOCS] Remove testenv annotations from doc snippet tests (#80023) 2021-11-05 18:38:50 -04:00
transform [DOCS] Add authorization info to get and update transform APIs (#87994) 2022-07-04 09:51:59 -07:00
troubleshooting Add troubleshooting guide for corrupt repository (#88391) 2022-07-14 13:37:02 +01:00
upgrade Docs for snapshots as simple archives (#86261) 2022-05-30 13:23:53 +02:00
vectors [DOCS] Remove testenv annotations from doc snippet tests (#80023) 2021-11-05 18:38:50 -04:00
aggregations.asciidoc Convert bucket aggs docs to runtime fields (#71202) 2021-04-02 12:12:06 -04:00
alias.asciidoc [DOCS] Fix default for is_write_index (#77006) (#77362) 2021-09-07 11:34:53 -04:00
analysis.asciidoc Update Lucene analysis base url (#84094) 2022-02-17 12:44:12 +01:00
api-conventions.asciidoc Fix a typo in api-conventions example (#88056) 2022-06-27 13:58:51 -04:00
cat.asciidoc [DOCS] Remove unneeded escapes 2021-04-26 12:14:45 -04:00
cluster.asciidoc How-to docs for increasing the total number of shards per node (#86214) 2022-05-10 09:13:27 +01:00
data-management.asciidoc reorder and merge data management and ILM doc pages (#84679) 2022-03-07 18:33:28 -05:00
data-rollup-transform.asciidoc [DOCS] Remove ifdefs for rollup refactor 2021-08-05 09:08:04 -04:00
datatiers.asciidoc [DOCS] Fix duplicate anchor (#85424) 2022-03-28 15:16:08 -07:00
dependencies-versions.asciidoc [DOCS] Added appendix to show dependencies (#67962) 2021-01-26 16:16:05 -08:00
docs.asciidoc [DOCS] Update single index APIs reference (#73103) 2021-05-14 11:53:34 -04:00
gs-index.asciidoc
high-availability.asciidoc [DOCS] Overhaul snapshot and restore docs (#79081) 2021-11-15 12:45:07 -05:00
how-to.asciidoc Move fix common cluster issues to troubleshooting (#87440) 2022-06-13 17:16:17 -07:00
index-extra-title-page.html [DOCS] Add index-extra-title-page.html for direct HTML migration (#50189) 2019-12-13 12:44:12 -05:00
index-modules.asciidoc Troubleshooting guides for disabled allocations (#86789) 2022-05-24 10:27:15 +01:00
index.asciidoc [DOCS] Remove ES quickstart. (#87939) 2022-06-23 14:25:48 -07:00
index.x.asciidoc [DOCS] Removes redundant index.asciidoc files (#30707) 2018-05-18 11:05:40 -07:00
indices.asciidoc Remove endpoint for freezing indices (#78918) 2021-10-26 06:37:56 -05:00
ingest.asciidoc Ingest: IngestDocument requires non-null version (#87665) 2022-06-15 07:50:45 -05:00
intro.asciidoc [DOCS] Update ES intro for stretched clusters (#77651) 2021-09-13 16:50:08 -04:00
links.asciidoc [DOCS] Rename ES Reference to ES Guide (#71198) 2021-04-01 15:38:41 -04:00
mapping.asciidoc Minor revision missed in merge. (#67282) 2021-01-11 13:50:06 -05:00
query-dsl.asciidoc Allow doc-values only search on geo_point fields (#83395) 2022-02-02 11:56:19 +01:00
redirects.asciidoc [DOCS] Adding discovery troubleshooting link in the master get help page (#87344) 2022-07-06 15:51:43 -04:00
release-notes.asciidoc [DOCS] Added 8.3.2 Release Notes stub (#88332) 2022-07-07 08:41:22 -05:00
scripting.asciidoc [DOCS] Add documentation for Painless field API (#83388) 2022-02-03 15:15:38 -05:00
search.asciidoc [DOCS] Add high-level guide for kNN search (#80857) 2021-11-30 14:17:39 -05:00
setup.asciidoc [DOCS] Overhaul snapshot and restore docs (#79081) 2021-11-15 12:45:07 -05:00
troubleshooting.asciidoc Add troubleshooting guide for corrupt repository (#88391) 2022-07-14 13:37:02 +01:00
upgrade.asciidoc Docs for snapshots as simple archives (#86261) 2022-05-30 13:23:53 +02:00