Fix Telegraf options removed in 1.40 on Linux and Windows (fieldpass -> fieldinclude, pid_tag -> tag_with)

Dev Guide.md

@@ -40,10 +40,11 @@ to preserve the overall 15-second metric-scrape limit; upstream's
 [1.40 documentation](https://github.com/influxdata/telegraf/blob/e9017dc3266369d6fa185e0e130af1d1d4021ce9/plugins/inputs/prometheus/README.md#L152-L157)
 explains that `response_timeout` now covers headers only, unlike the
 [1.24.2 client timeout](https://github.com/influxdata/telegraf/blob/9550e7a533dd00632e14435e87ed3eb4b04832c6/plugins/inputs/prometheus/prometheus.go#L254-L261).
-Procstat uses `tag_with = ["pid"]` to retain PID tags. Existing `fieldpass`/`fielddrop`
-names are unchanged on both OSes because
-[1.40 still parses them](https://github.com/influxdata/telegraf/blob/e9017dc3266369d6fa185e0e130af1d1d4021ce9/config/config.go#L1649-L1687).
-Linux rendering remains unchanged. Run `ruby build/common/installer/scripts/tomlparser-prom-customconfig_test.rb`
+Procstat uses `tag_with = ["pid"]` to retain PID tags. Both OSes render
+`fieldinclude`/`fieldexclude` instead of the deprecated `fieldpass`/`fielddrop`
+Telegraf options. The public ConfigMap keys remain `fieldpass`/`fielddrop` for
+backward compatibility; the shared parser maps them to the current Telegraf names.
+Run `ruby build/common/installer/scripts/tomlparser-prom-customconfig_test.rb`
 for rendering coverage with and without namespace filters. On Windows, set
 `TELEGRAF_WINDOWS_BINARY` to the extracted `telegraf.exe` to also load the generated
 configs with that binary in bounded `--test` mode, without Kubernetes access or

build/common/installer/scripts/tomlparser-prom-customconfig.rb

@@ -186,8 +186,6 @@ def createPrometheusPluginsWithNamespaceSetting(monitorKubernetesPods, monitorKu
     new_contents = new_contents.gsub("$AZMON_TELEGRAF_CUSTOM_PROM_KUBERNETES_FIELD_SELECTOR", "# Commenting this out since new plugins will be created per namespace\n  # $AZMON_TELEGRAF_CUSTOM_PROM_KUBERNETES_FIELD_SELECTOR")
     new_contents = new_contents.gsub("$AZMON_TELEGRAF_CUSTOM_PROM_SCRAPE_SCOPE", "# Commenting this out since new plugins will be created per namespace\n  # $AZMON_TELEGRAF_CUSTOM_PROM_SCRAPE_SCOPE")
 
-    timeout_config_key = "timeout"
-
     pluginConfigsWithNamespaces = ""
     podScrapeScope = (@controller.casecmp(@replicaset) == 0) ? "cluster" : "node"
     monitorKubernetesPodsNamespaces.each do |namespace|
@@ -208,11 +206,11 @@ def createPrometheusPluginsWithNamespaceSetting(monitorKubernetesPods, monitorKu
   monitor_kubernetes_pods_namespace = #{toTomlBasicString(namespace)}
   kubernetes_label_selector = #{toTomlBasicString(kubernetesLabelSelectors)}
   kubernetes_field_selector = #{toTomlBasicString(kubernetesFieldSelectors)}
-  fieldpass = #{fieldPassSetting}
-  fielddrop = #{fieldDropSetting}
+  fieldinclude = #{fieldPassSetting}
+  fieldexclude = #{fieldDropSetting}
   metric_version = #{@metricVersion}
   url_tag = #{toTomlBasicString(@urlTag)}
-  #{timeout_config_key} = #{toTomlBasicString(@responseTimeout)}
+  timeout = #{toTomlBasicString(@responseTimeout)}
   tls_ca = #{toTomlBasicString(@tlsCa)}
   insecure_skip_verify = #{@insecureSkipVerify}\n"
         end

build/common/installer/scripts/tomlparser-prom-customconfig_test.rb

@@ -255,19 +255,24 @@ def test_kubernetes_namespace_validation
       result = run_parser(scenario, configmap_for(scenario, "fieldpass = ['''#{BREAKOUT}''']"))
       assert_no_injected_plugins(scenario, result[:conf], "the #{scenario} fieldpass array")
       # The value survives, but only as a single escaped string.
-      assert_includes collect_values(result[:conf], "fieldpass"), BREAKOUT
+      assert_includes collect_values(result[:conf], "fieldinclude"), BREAKOUT
     end
 
     define_method("test_fielddrop_breakout_is_neutralized_#{scenario}") do
       result = run_parser(scenario, configmap_for(scenario, "fielddrop = ['''#{BREAKOUT}''']"))
       assert_no_injected_plugins(scenario, result[:conf], "the #{scenario} fielddrop array")
-      assert_includes collect_values(result[:conf], "fielddrop"), BREAKOUT
+      assert_includes collect_values(result[:conf], "fieldexclude"), BREAKOUT
     end
 
     define_method("test_valid_settings_are_preserved_#{scenario}") do
-      result = run_parser(scenario, configmap_for(scenario, "interval = \"45s\"\nfieldpass = [\"a\",\"b\"]"))
+      result = run_parser(scenario, configmap_for(scenario, "interval = \"45s\"\nfieldpass = [\"a\",\"b\"]\nfielddrop = [\"c\"]"))
       assert_includes result[:conf], "interval = \"45s\"", "a valid interval must be preserved"
-      assert_includes result[:conf], "fieldpass = [\"a\",\"b\"]", "array formatting must be unchanged"
+      assert_includes result[:conf], "fieldinclude = [\"a\",\"b\"]", "array formatting must be unchanged"
+      assert_includes result[:conf], "fieldexclude = [\"c\"]", "array formatting must be unchanged"
+      parse_generated_toml(result[:conf])["inputs"]["prometheus"].each do |plugin|
+        refute plugin.key?("fieldpass")
+        refute plugin.key?("fielddrop")
+      end
       assert_no_injected_plugins(scenario, result[:conf], "a benign #{scenario} configuration")
     end
   end
@@ -333,8 +338,10 @@ def test_namespace_plugins_are_generated_for_valid_namespaces
           assert_equal "scrapeUrl", plugin["url_tag"]
           assert_equal "pod_namespace", plugin["pod_namespace_label_name"]
           assert_equal (scenario == :replicaset ? "cluster" : "node"), plugin["pod_scrape_scope"]
-          assert_equal ["requests_total"], plugin["fieldpass"]
-          assert_equal ["debug_total"], plugin["fielddrop"]
+          assert_equal ["requests_total"], plugin["fieldinclude"]
+          assert_equal ["debug_total"], plugin["fieldexclude"]
+          refute plugin.key?("fieldpass")
+          refute plugin.key?("fielddrop")
           assert_equal "app=metrics", plugin["kubernetes_label_selector"]
           assert_equal "spec.nodeName=test-node", plugin["kubernetes_field_selector"]
         end
@@ -375,6 +382,7 @@ def test_windows_rendered_configs_load_in_packaged_telegraf
           output = readers.map(&:value).join("\n")
           assert completed, "Telegraf config smoke test exceeded 20 seconds"
           assert process.value.success?, "Telegraf rejected the rendered config for #{namespaces.inspect}: #{output}"
+          refute_match(/Deprecation(?:Error|Warning).*"(?:fieldpass|fielddrop|pid_tag)"/, output)
         end
       end
     end
@@ -388,7 +396,8 @@ def test_windows_process_metrics_config_preserves_pid_tags_and_fields
     config["inputs"]["procstat"].each do |plugin|
       assert_equal ["pid"], plugin["tag_with"]
       refute plugin.key?("pid_tag")
-      assert_equal ["cpu_usage", "memory_rss"], plugin["fieldpass"]
+      assert_equal ["cpu_usage", "memory_rss"], plugin["fieldinclude"]
+      refute plugin.key?("fieldpass")
       assert_equal "native", plugin["pid_finder"]
       assert_equal "agent_telemetry", plugin["name_override"]
       assert_equal "t.azm.ms/", plugin["name_prefix"]
@@ -400,6 +409,51 @@ def test_windows_process_metrics_config_preserves_pid_tags_and_fields
     )
   end
 
+  def test_linux_process_metrics_config_preserves_pid_tags_and_fields
+    path = File.join(REPO_ROOT, "build/linux/installer/conf/telegraf-ama-logs-process-metrics.conf")
+    config = Tomlrb.load_file(path)
+    assert_equal({ "telegraf_role" => "ama-logs-process-metrics" }, config["global_tags"])
+    assert_equal 9, config["inputs"]["procstat"].length
+    config["inputs"]["procstat"].each do |plugin|
+      assert_equal ["pid"], plugin["tag_with"]
+      refute plugin.key?("pid_tag")
+      assert_equal ["cpu_usage", "memory_rss"], plugin["fieldinclude"]
+      refute plugin.key?("fieldpass")
+      assert_equal "native", plugin["pid_finder"]
+      assert_equal "agent_telemetry", plugin["name_override"]
+      assert_equal "t.azm.ms/", plugin["name_prefix"]
+      assert_equal "$CONTROLLER_TYPE", plugin["tags"]["ControllerType"]
+      # ORDER matters: renamed keys must stay above [inputs.procstat.tags], otherwise
+      # TOML binds them to the nested table and telegraf never sees them.
+      ["tag_with", "fieldinclude"].each do |key|
+        refute plugin["tags"].key?(key), "#{key} must not be absorbed by [inputs.procstat.tags]"
+      end
+    end
+    assert_equal(
+      { "ai.cloud.role" => "ControllerType", "ai.cloud.roleInstance" => "PodName" },
+      config["outputs"]["application_insights"].first["context_tag_sources"]
+    )
+  end
+
+  # Telegraf treats an unrecognized plugin option as a fatal config-load error, so a single
+  # reintroduced legacy name silently stops the agent from loading that configuration.
+  REMOVED_TELEGRAF_OPTIONS = %w[fieldpass fielddrop pid_tag ignore_protocol_stats].freeze
+
+  def test_shipped_telegraf_templates_do_not_use_removed_options
+    templates = Dir.glob(File.join(REPO_ROOT, "build/{linux,windows}/installer/conf/telegraf*.conf")).sort
+    refute_empty templates, "expected to find shipped telegraf templates"
+
+    templates.each do |path|
+      offenders = File.readlines(path).each_with_index.filter_map do |line, index|
+        stripped = line.strip
+        next if stripped.empty? || stripped.start_with?("#")
+        option = REMOVED_TELEGRAF_OPTIONS.find { |name| stripped =~ /\A#{name}\s*=/ }
+        "#{File.basename(path)}:#{index + 1}: #{stripped}" if option
+      end
+      assert_empty offenders, "removed telegraf options must not be used in shipped templates"
+    end
+  end
+
   def test_selectors_are_escaped_in_generated_namespace_plugins
     body = "monitor_kubernetes_pods = true\n" \
            "monitor_kubernetes_pods_namespaces = [\"default\"]\n" \
@@ -411,6 +465,28 @@ def test_selectors_are_escaped_in_generated_namespace_plugins
     assert_equal BREAKOUT, plugin["kubernetes_label_selector"]
   end
 
+  def test_namespace_filters_preserve_modern_keys_and_escaping
+    [:replicaset, :sidecar, :windows].each do |scenario|
+      body = "monitor_kubernetes_pods = true\n" \
+             "monitor_kubernetes_pods_namespaces = [\"default\"]\n" \
+             "fieldpass = ['''#{BREAKOUT}''']\n" \
+             "fielddrop = ['''#{BREAKOUT}''']\n" \
+             "kubernetes_label_selector = '''#{BREAKOUT}'''\n" \
+             "kubernetes_field_selector = '''#{BREAKOUT}'''"
+      result = run_parser(scenario, configmap_for(scenario, body))
+      assert_no_injected_plugins(scenario, result[:conf], "namespaced #{scenario} filters")
+      plugin = parse_generated_toml(result[:conf])["inputs"]["prometheus"]
+               .find { |entry| entry["monitor_kubernetes_pods_namespace"] == "default" }
+      refute_nil plugin
+      assert_equal [BREAKOUT], plugin["fieldinclude"]
+      assert_equal [BREAKOUT], plugin["fieldexclude"]
+      assert_equal BREAKOUT, plugin["kubernetes_label_selector"]
+      assert_equal BREAKOUT, plugin["kubernetes_field_selector"]
+      assert_equal(scenario == :replicaset ? "cluster" : "node", plugin["pod_scrape_scope"])
+      ["fieldpass", "fielddrop"].each { |key| refute plugin.key?(key), "#{scenario} must not emit #{key}" }
+    end
+  end
+
   def monitored_namespaces(conf)
     parse_generated_toml(conf)["inputs"]["prometheus"]
       .map { |plugin| plugin["monitor_kubernetes_pods_namespace"] }

build/linux/installer/conf/telegraf-ama-logs-process-metrics.conf

@@ -22,9 +22,9 @@
   exe = "mdsd"
   interval = "60s"
   pid_finder = "native"
-  pid_tag = true
+  tag_with = ["pid"]
   name_override = "agent_telemetry"
-  fieldpass = ["cpu_usage", "memory_rss"]
+  fieldinclude = ["cpu_usage", "memory_rss"]
   [inputs.procstat.tags]
     Computer = "placeholder_hostname"
     PodName = "placeholder_podname"
@@ -38,9 +38,9 @@
   exe = "fluent-bit"
   interval = "60s"
   pid_finder = "native"
-  pid_tag = true
+  tag_with = ["pid"]
   name_override = "agent_telemetry"
-  fieldpass = ["cpu_usage", "memory_rss"]
+  fieldinclude = ["cpu_usage", "memory_rss"]
   [inputs.procstat.tags]
     Computer = "placeholder_hostname"
     PodName = "placeholder_podname"
@@ -54,9 +54,9 @@
   pattern = "telegraf(-rs|-prom-side-car)?.conf"
   interval = "60s"
   pid_finder = "native"
-  pid_tag = true
+  tag_with = ["pid"]
   name_override = "agent_telemetry"
-  fieldpass = ["cpu_usage", "memory_rss"]
+  fieldinclude = ["cpu_usage", "memory_rss"]
   [inputs.procstat.tags]
     Computer = "placeholder_hostname"
     PodName = "placeholder_podname"
@@ -70,9 +70,9 @@
   pattern = "telegraf-ama-logs-process-metrics.conf"
   interval = "60s"
   pid_finder = "native"
-  pid_tag = true
+  tag_with = ["pid"]
   name_override = "agent_telemetry"
-  fieldpass = ["cpu_usage", "memory_rss"]
+  fieldinclude = ["cpu_usage", "memory_rss"]
   [inputs.procstat.tags]
     Computer = "placeholder_hostname"
     PodName = "placeholder_podname"
@@ -85,9 +85,9 @@
   name_prefix = "t.azm.ms/"
   interval = "60s"
   pid_finder = "native"
-  pid_tag = true
+  tag_with = ["pid"]
   name_override = "agent_telemetry"
-  fieldpass = ["cpu_usage", "memory_rss"]
+  fieldinclude = ["cpu_usage", "memory_rss"]
   [inputs.procstat.tags]
     Computer = "placeholder_hostname"
     PodName = "placeholder_podname"
@@ -104,9 +104,9 @@
   exe = "crond"
   interval = "60s"
   pid_finder = "native"
-  pid_tag = true
+  tag_with = ["pid"]
   name_override = "agent_telemetry"
-  fieldpass = ["cpu_usage", "memory_rss"]
+  fieldinclude = ["cpu_usage", "memory_rss"]
   [inputs.procstat.tags]
     Computer = "placeholder_hostname"
     PodName = "placeholder_podname"
@@ -120,9 +120,9 @@
   exe = "inotifywait"
   interval = "60s"
   pid_finder = "native"
-  pid_tag = true
+  tag_with = ["pid"]
   name_override = "agent_telemetry"
-  fieldpass = ["cpu_usage", "memory_rss"]
+  fieldinclude = ["cpu_usage", "memory_rss"]
   [inputs.procstat.tags]
     Computer = "placeholder_hostname"
     PodName = "placeholder_podname"
@@ -135,9 +135,9 @@
   name_prefix = "t.azm.ms/"
   interval = "60s"
   pid_finder = "native"
-  pid_tag = true
+  tag_with = ["pid"]
   name_override = "agent_telemetry"
-  fieldpass = ["cpu_usage", "memory_rss"]
+  fieldinclude = ["cpu_usage", "memory_rss"]
   [inputs.procstat.tags]
     Computer = "placeholder_hostname"
     PodName = "placeholder_podname"
@@ -154,9 +154,9 @@
   pattern = "main.sh"
   interval = "60s"
   pid_finder = "native"
-  pid_tag = true
+  tag_with = ["pid"]
   name_override = "agent_telemetry"
-  fieldpass = ["cpu_usage", "memory_rss"]
+  fieldinclude = ["cpu_usage", "memory_rss"]
   [inputs.procstat.tags]
     Computer = "placeholder_hostname"
     PodName = "placeholder_podname"

build/linux/installer/conf/telegraf-prom-side-car.conf

@@ -183,8 +183,8 @@
   $AZMON_TELEGRAF_CUSTOM_PROM_KUBERNETES_LABEL_SELECTOR
   $AZMON_TELEGRAF_CUSTOM_PROM_KUBERNETES_FIELD_SELECTOR
 
-  fieldpass = $AZMON_TELEGRAF_CUSTOM_PROM_FIELDPASS
-  fielddrop = $AZMON_TELEGRAF_CUSTOM_PROM_FIELDDROP
+  fieldinclude = $AZMON_TELEGRAF_CUSTOM_PROM_FIELDPASS
+  fieldexclude = $AZMON_TELEGRAF_CUSTOM_PROM_FIELDDROP
 
   metric_version = 2
   url_tag = "scrapeUrl"

build/linux/installer/conf/telegraf-rs.conf

@@ -574,8 +574,8 @@
   $AZMON_TELEGRAF_CUSTOM_PROM_KUBERNETES_LABEL_SELECTOR
   $AZMON_TELEGRAF_CUSTOM_PROM_KUBERNETES_FIELD_SELECTOR
 
-  fieldpass = $AZMON_TELEGRAF_CUSTOM_PROM_FIELDPASS
-  fielddrop = $AZMON_TELEGRAF_CUSTOM_PROM_FIELDDROP
+  fieldinclude = $AZMON_TELEGRAF_CUSTOM_PROM_FIELDPASS
+  fieldexclude = $AZMON_TELEGRAF_CUSTOM_PROM_FIELDDROP
 
   metric_version = 2
   url_tag = "scrapeUrl"
@@ -641,7 +641,7 @@ $AZMON_TELEGRAF_OSM_PROM_PLUGINS
   #name_prefix="container.azm.ms/"
   ## An array of urls to scrape metrics from.
   urls = $AZMON_INTEGRATION_NPM_METRICS_URL_LIST_CLUSTER
-  fielddrop = $AZMON_INTEGRATION_NPM_METRICS_DROP_LIST_CLUSTER
+  fieldexclude = $AZMON_INTEGRATION_NPM_METRICS_DROP_LIST_CLUSTER
 
   metric_version = 2
   url_tag = "scrapeUrl"

build/linux/installer/conf/telegraf.conf

@@ -448,7 +448,7 @@
 
   ## Ignore mount points by filesystem type.
   ignore_fs = ["tmpfs", "devtmpfs", "devfs", "overlay", "aufs", "squashfs", "cifs", "fuse"]
-  fieldpass = ["free", "used", "used_percent"]
+  fieldinclude = ["free", "used", "used_percent"]
   taginclude = ["device","path","hostName"]
   # Below due to Bug - https://github.com/influxdata/telegraf/issues/5615
   # ORDER matters here!! - i.e the below should be the LAST modifier
@@ -489,7 +489,7 @@
   ## The typical use case is for LVM volumes, to get the VG/LV name instead of
   ## the near-meaningless DM-0 name.
   # name_templates = ["$ID_FS_LABEL","$DM_VG_NAME/$DM_LV_NAME"]
-  fieldpass = ["reads", "read_bytes", "read_time", "writes", "write_bytes", "write_time", "io_time", "iops_in_progress"]
+  fieldinclude = ["reads", "read_bytes", "read_time", "writes", "write_bytes", "write_time", "io_time", "iops_in_progress"]
   taginclude = ["name","hostName"]
 
 # Read metrics about network interface usage
@@ -501,12 +501,11 @@
   ##
   # interfaces = ["eth0"]
   ##
-  ## On linux systems telegraf also collects protocol stats.
-  ## Setting ignore_protocol_stats to true will skip reporting of protocol metrics.
+  ## NOTE: protocol stats collection was removed in telegraf 1.37.0 and the
+  ## ignore_protocol_stats option became a no-op (removed in 1.45.0), so it is
+  ## no longer set here. Current behavior matches the previous "true" setting.
   ##
-  ignore_protocol_stats = true
-  ##
-  fieldpass = ["bytes_recv", "bytes_sent", "err_in", "err_out"]
+  fieldinclude = ["bytes_recv", "bytes_sent", "err_in", "err_out"]
   taginclude = ["interface","hostName"]
 
 # net plugin is collecting loopback interface metrics by default from telegraf 1.34.3
@@ -652,7 +651,7 @@
   name_prefix="container.azm.ms/"
   ## An array of urls to scrape metrics from.
   urls = ["$CADVISOR_METRICS_URL"]
-  fieldpass = ["$KUBELET_RUNTIME_OPERATIONS_METRIC", "$KUBELET_RUNTIME_OPERATIONS_ERRORS_METRIC", "$KUBELET_RUNTIME_OPERATIONS_TOTAL_METRIC", "$KUBELET_RUNTIME_OPERATIONS_ERRORS_TOTAL_METRIC"]
+  fieldinclude = ["$KUBELET_RUNTIME_OPERATIONS_METRIC", "$KUBELET_RUNTIME_OPERATIONS_ERRORS_METRIC", "$KUBELET_RUNTIME_OPERATIONS_TOTAL_METRIC", "$KUBELET_RUNTIME_OPERATIONS_ERRORS_TOTAL_METRIC"]
 
   metric_version = 2
   url_tag = "scrapeUrl"
@@ -697,7 +696,7 @@
 
   # <= 1.18: metric name is kubelet_running_pod_count
   # >= 1.19: metric name changed to kubelet_running_pods
-  fieldpass = ["kubelet_running_pod_count","kubelet_running_pods","volume_manager_total_volumes", "kubelet_node_config_error", "process_resident_memory_bytes", "process_cpu_seconds_total"]
+  fieldinclude = ["kubelet_running_pod_count","kubelet_running_pods","volume_manager_total_volumes", "kubelet_node_config_error", "process_resident_memory_bytes", "process_cpu_seconds_total"]
 
   metric_version = 2
   url_tag = "scrapeUrl"
@@ -721,9 +720,9 @@
   ## An array of urls to scrape metrics from.
   urls = $AZMON_DS_PROM_URLS
 
-  fieldpass = $AZMON_DS_PROM_FIELDPASS
+  fieldinclude = $AZMON_DS_PROM_FIELDPASS
 
-  fielddrop = $AZMON_DS_PROM_FIELDDROP
+  fieldexclude = $AZMON_DS_PROM_FIELDDROP
 
   metric_version = 2
   url_tag = "scrapeUrl"

build/linux/installer/scripts/tomlparser-osm-config.rb

@@ -108,7 +108,7 @@ def replaceOsmTelegrafConfigPlaceHolders
   monitor_kubernetes_pods = true
   pod_scrape_scope = \"#{(@controllerType.casecmp(@replicaset) == 0) ? "cluster" : "node"}\"
   monitor_kubernetes_pods_namespace = \"#{namespace}\"
-  fieldpass = #{@fieldPassSetting}
+  fieldinclude = #{@fieldPassSetting}
   metric_version = #{@metricVersion}
   url_tag = \"#{@urlTag}\"
   bearer_token = \"#{@bearerToken}\"