From 4046187d95940dc3b19de81ae237cff6b087ec00 Mon Sep 17 00:00:00 2001
From: Toan Nguyen <hgiasac@gmail.com>
Date: Wed, 18 Dec 2024 22:46:35 +0700
Subject: [PATCH] docs: add more OpenTelemetry collector configuration examples

PR-URL: https://github.com/hasura/graphql-engine-mono/pull/11116
GitOrigin-RevId: 7b341b9ae84b0eb90ad87ac0878020fb0b6f9f88
---
 .../observability/enterprise/dotenv           |  2 +-
 .../otel-collector/otel-collector-config.yaml | 63 ++++++++++++++++---
 .../opentelemetry/best-practices.mdx          |  2 +
 3 files changed, 57 insertions(+), 10 deletions(-)

diff --git a/community/boilerplates/observability/enterprise/dotenv b/community/boilerplates/observability/enterprise/dotenv
index 01fe5bb3a7c3c..5c5f1a58df180 100644
--- a/community/boilerplates/observability/enterprise/dotenv
+++ b/community/boilerplates/observability/enterprise/dotenv
@@ -1,5 +1,5 @@
 # HGE
-HGE_IMAGE=hasura/graphql-engine:v2.40.0.cli-migrations-v3
+HGE_IMAGE=hasura/graphql-engine:v2.45.1.cli-migrations-v3
 HGE_PORT=8080
 HGE_PRO_KEY=
 HGE_EE_LICENSE_KEY=
diff --git a/community/boilerplates/observability/enterprise/otel-collector/otel-collector-config.yaml b/community/boilerplates/observability/enterprise/otel-collector/otel-collector-config.yaml
index 1b13a48dc6382..6d380f1d4a780 100644
--- a/community/boilerplates/observability/enterprise/otel-collector/otel-collector-config.yaml
+++ b/community/boilerplates/observability/enterprise/otel-collector/otel-collector-config.yaml
@@ -22,10 +22,10 @@ exporters:
   #     key: ${DD_API_KEY}
 
   ## New Relic exporter
-  otlp/newrelic:
-    endpoint: otlp.nr-data.net:4317
-    headers:
-      api-key: ${NEW_RELIC_API_KEY}
+  # otlp/newrelic:
+  #   endpoint: otlp.nr-data.net:4317
+  #   headers:
+  #     api-key: ${NEW_RELIC_API_KEY}
 
   ## Export traces to Jaeger via OTLP protocol.
   ## You can also export traces directly to jaeger to graphql-engine.
@@ -51,8 +51,12 @@ exporters:
   ## download and rename to gcp-sa.json so the otel-collector can mount the service account
   ##
   # googlecloud:
+  #   trace:
+  #   metric:
+  #     compression: gzip
   #   log:
   #     default_log_name: example-otel.io/hasura-exported-log
+  #     compression: gzip
 
 processors:
   batch:
@@ -113,9 +117,50 @@ processors:
 
     logs:
       log_record:
-        # You can filter unused logs to save cost, or configure HASURA_GRAPHQL_ENABLED_LOG_TYPES in GraphQL Engine to disable unused log types
-        - 'attributes["type"] == "query-log" and IsMatch(body["query"]["operationName"], "UnknownQuery")'
-        - 'attributes["type"] == "http-log" and IsMatch(body["operation"]["query"]["operationName"], "UnknownQuery")'
+        ## Disable debug logs. You can also set HASURA_GRAPHQL_LOG_LEVEL=info.
+        - severity_number == SEVERITY_NUMBER_DEBUG
+
+        ## Deprecated log types. You should export OTEL traces and metrics instead.
+        - 'attributes["type"] == "metrics" or attributes["type"] == "tracing-log" or attributes["type"] == "livequery-poller-log"'
+
+        ## http-log and websocket-log already include most of information in query-log.
+        ## You can ignore this log type unless you want to debug generated SQL.
+        # - 'attributes["type"] == "query-log"'
+
+        ## You can filter unused logs to save cost, or configure HASURA_GRAPHQL_ENABLED_LOG_TYPES in GraphQL Engine to disable unused log types
+        # - 'attributes["type"] == "query-log" and IsMatch(body["query"]["operationName"], "UnknownQuery")'
+        # - 'attributes["type"] == "http-log" and IsMatch(body["operation"]["query"]["operationName"], "UnknownQuery")'
+
+  transform/logs:
+    error_mode: ignore
+    log_statements:
+      ## Remove the transform context in trigger logs
+      # - context: log
+      #   conditions:
+      #     - 'attributes["type"] == "scheduled-trigger" or attributes["type"] == "event-trigger"'
+      #   statements:
+      #     - 'delete_key(body["request"], "req_transform_ctx")'
+
+      ## Truncate very long query strings in http-log
+      - context: log
+        conditions:
+          - 'attributes["type"] == "http-log" and IsMap(body["operation"])'
+        statements:
+          - 'truncate_all(body["operation"], 4096)'
+
+      ## Truncate very long query strings in http-log
+      - context: log
+        conditions:
+          - 'attributes["type"] == "websocket-log" and IsMap(body["event.detail.query"])'
+        statements:
+          - 'truncate_all(body["event.detail.query"], 4096)'
+
+      ## Truncate very long generated_sql strings in query-log
+      - context: log
+        conditions:
+          - 'attributes["type"] == "query-log" and IsMap(body["generated_sql"])'
+        statements:
+          - 'truncate_all(body["generated_sql"], 4096)'
 
 service:
   pipelines:
@@ -126,8 +171,8 @@ service:
 
     logs:
       receivers: [otlp]
-      processors: [filter, memory_limiter, batch]
-      exporters: [loki]
+      processors: [filter, transform/logs, memory_limiter, batch]
+      exporters: [loki, debug]
 
     metrics:
       receivers: [otlp]
diff --git a/docs/docs/observability/opentelemetry/best-practices.mdx b/docs/docs/observability/opentelemetry/best-practices.mdx
index 43cb417bf82f7..fe7b06d9f3034 100644
--- a/docs/docs/observability/opentelemetry/best-practices.mdx
+++ b/docs/docs/observability/opentelemetry/best-practices.mdx
@@ -85,6 +85,8 @@ processors:
         - 'attributes["type"] == "http-log" and IsMatch(body["operation"]["query"]["operationName"], "UnknownQuery")'
 ```
 
+Check out more configuration examples [here](https://github.com/hasura/graphql-engine/blob/master/community/boilerplates/observability/enterprise/otel-collector/otel-collector-config.yaml).
+
 ## Monitoring
 
 ### Logging