From 362fe84b70be88431f11f39e1e58eba79183a1d3 Mon Sep 17 00:00:00 2001 From: Colleen McGinnis Date: Wed, 18 Dec 2024 15:11:50 -0600 Subject: [PATCH] restructure custom filters page (#4703) (cherry picked from commit 487c3b53138eedf50b715e5c65093514789b0668) --- .../data-security/custom-filter.asciidoc | 96 ++++++++++++------- .../apm/security/data-security/index.asciidoc | 2 - 2 files changed, 61 insertions(+), 37 deletions(-) diff --git a/docs/en/observability/apm/security/data-security/custom-filter.asciidoc b/docs/en/observability/apm/security/data-security/custom-filter.asciidoc index 300f5ef6c5..ad04fc1d7b 100644 --- a/docs/en/observability/apm/security/data-security/custom-filter.asciidoc +++ b/docs/en/observability/apm/security/data-security/custom-filter.asciidoc @@ -1,19 +1,48 @@ [[apm-custom-filter]] = Custom filters -include::{observability-docs-root}/docs/en/observability/apm/security/data-security/index.asciidoc[tag=custom-filters] +Custom filters, including <> and +<>, allow you to filter or redact APM data on ingestion. [discrete] [[apm-filters-ingest-pipeline]] -== Create an ingest pipeline filter +== Ingest pipeline filters -Ingest node pipelines specify a series of processors that transform data in a specific way. +Ingest pipelines specify a series of processors that transform data in a specific way. Transformation happens prior to indexing--inflicting no performance overhead on the monitored application. Pipelines are a flexible and easy way to filter or obfuscate Elastic APM data. +Features of this approach: + +* Filters are applied at ingestion time. +* All Elastic APM agents and fields are supported. +* Data leaves the instrumented service. +* There are no performance overhead implications on the instrumented service. + +For a step-by-step example, refer to <>. + +[discrete] +[[apm-filters-in-agent]] +== APM agent filters + +Some APM agents offer a way to manipulate or drop APM events _before_ they are sent to APM Server. + +Features of this approach: + +* Data is sanitized before leaving the instrumented service. +* Not supported by all Elastic APM agents. +* Potential overhead implications on the instrumented service. + +Refer to the relevant agent's documentation for more information and examples: + +* .NET: {apm-dotnet-ref-v}/public-api.html#filter-api[Filter API]. +* Node.js: {apm-node-ref-v}/agent-api.html#apm-add-filter[`addFilter()`]. +* Python: {apm-py-ref-v}/sanitizing-data.html[custom processors]. +* Ruby: {apm-ruby-ref-v}/api.html#api-agent-add-filter[`add_filter()`]. + [discrete] [[apm-filters-ingest-pipeline-tutorial]] -=== Tutorial: redact sensitive information +== Tutorial: Use an ingest pipeline to redact sensitive information Say you decide to <> but quickly notice that sensitive information is being collected in the @@ -27,10 +56,16 @@ but quickly notice that sensitive information is being collected in the } ---- -**Create a pipeline** +To obfuscate the passwords stored in the request body, you can use a series of {ref}/processors.html[ingest processors]. + +[discrete] +=== Create a pipeline + +TIP: This tutorial uses the {ref}/ingest-apis.html[Ingest APIs], but it's also possible +to create a pipeline using the UI. In Kibana, go to *Stack Management* → +*Ingest Pipelines* → *Create pipeline* → *New pipeline* or use the +{kibana-ref}/introduction.html#kibana-navigation-search[global search field]. -To obfuscate the passwords stored in the request body, -you can use a series of {ref}/processors.html[ingest processors]. To start, create a pipeline with a simple description and an empty array of processors: [source,json] @@ -44,7 +79,8 @@ To start, create a pipeline with a simple description and an empty array of proc ---- <1> The processors defined below will go in this array -**Add a JSON processor** +[discrete] +==== Add a JSON processor Add your first processor to the processors array. Because the agent captures the request body as a string, use the @@ -62,7 +98,8 @@ Save this JSON object in a new field: } ---- -**Add a set processor** +[discrete] +==== Add a set processor If `body.original_json` is not `null`, i.e., it exists, we'll redact the `password` with the {ref}/set-processor.html[set processor], by setting the value of `body.original_json.password` to `"redacted"`: @@ -78,7 +115,8 @@ by setting the value of `body.original_json.password` to `"redacted"`: } ---- -**Add a convert processor** +[discrete] +==== Add a convert processor Use the {ref}/convert-processor.html[convert processor] to convert the JSON value of `body.original_json` to a string and set it as the `body.original` value: @@ -95,7 +133,8 @@ Use the {ref}/convert-processor.html[convert processor] to convert the JSON valu } ---- -**Add a remove processor** +[discrete] +==== Add a remove processor Finally, use the {ref}/remove-processor.html[remove processor] to remove the `body.original_json` field: @@ -103,17 +142,18 @@ Finally, use the {ref}/remove-processor.html[remove processor] to remove the `bo ---- { "remove": { - "field": "http.request.body.original", + "field": "http.request.body.original_json", "if": "ctx?.http?.request?.body?.original_json != null", "ignore_failure": true } } ---- -**Register the pipeline** +[discrete] +==== Register the pipeline -Now we'll put it all together. -Use the {ref}/put-pipeline-api.html[create or update pipeline API] to register the new pipeline in {es}. +Then put it all together, +and use the {ref}/put-pipeline-api.html[create or update pipeline API] to register the new pipeline in {es}. Name the pipeline `apm_redacted_body_password`: [source,console] @@ -156,7 +196,8 @@ PUT _ingest/pipeline/apm_redacted_body_password } ---- -**Test the pipeline** +[discrete] +=== Test the pipeline Prior to enabling this new pipeline, you can test it with the {ref}/simulate-pipeline-api.html[simulate pipeline API]. This API allows you to run multiple documents through a pipeline to ensure it is working correctly. @@ -251,7 +292,8 @@ The API response should be similar to this: As expected, only the first simulated document has a redacted password field. All other documents are unaffected. -**Create an `@custom` pipeline** +[discrete] +=== Create a `@custom` pipeline The final step in this process is to call the newly created `apm_redacted_body_password` pipeline from the `@custom` pipeline of the data stream you wish to edit. @@ -276,25 +318,9 @@ PUT _ingest/pipeline/traces-apm@custom ---- <1> The name of the pipeline we previously created -TIP: If you prefer using a GUI, you can instead use the **Ingest Pipelines** page in {kib}. -To open **Ingest Pipelines**, find **Stack Management** in the main menu or use the {kibana-ref}/introduction.html#kibana-navigation-search[global search field]. -Click **Create pipeline** and use the same naming convention explained previously to ensure your new pipeline matches the correct APM data stream. - That's it! Passwords will now be redacted from your APM HTTP body data. -To learn more about ingest pipelines, see <>. - [discrete] -[[apm-filters-in-agent]] -== APM agent filters - -Some APM agents offer a way to manipulate or drop APM events _before_ they are sent to the APM Server. -Please see the relevant agent's documentation for more information and examples: +=== Next steps -// * Go: {apm-go-ref-v}/[] -// * Java: {apm-java-ref-v}/[] -* .NET: {apm-dotnet-ref-v}/public-api.html#filter-api[Filter API]. -* Node.js: {apm-node-ref-v}/agent-api.html#apm-add-filter[`addFilter()`]. -// * PHP: {apm-php-ref-v}[] -* Python: {apm-py-ref-v}/sanitizing-data.html[custom processors]. -* Ruby: {apm-ruby-ref-v}/api.html#api-agent-add-filter[`add_filter()`]. +To learn more about ingest pipelines, see <>. diff --git a/docs/en/observability/apm/security/data-security/index.asciidoc b/docs/en/observability/apm/security/data-security/index.asciidoc index 0b041e6c6e..a96ffe86a2 100644 --- a/docs/en/observability/apm/security/data-security/index.asciidoc +++ b/docs/en/observability/apm/security/data-security/index.asciidoc @@ -42,7 +42,6 @@ Built-in data filters allow you to filter or turn off ingestion of the following [[apm-custom-data-filters]] == Custom filters -// tag::custom-filters[] Custom filters allow you to filter or redact other types of APM data on ingestion: |==== @@ -54,7 +53,6 @@ There are no performance overhead implications on the instrumented service. Data is sanitized before leaving the instrumented service. Potential overhead implications on the instrumented service |==== -// end::custom-filters[] [float] [[apm-sensitive-fields]]