consul/agent/hcp/telemetry_provider.go

323 lines
9.4 KiB
Go
Raw Normal View History

[COMPLIANCE] License changes (#18443) * Adding explicit MPL license for sub-package This directory and its subdirectories (packages) contain files licensed with the MPLv2 `LICENSE` file in this directory and are intentionally licensed separately from the BSL `LICENSE` file at the root of this repository. * Adding explicit MPL license for sub-package This directory and its subdirectories (packages) contain files licensed with the MPLv2 `LICENSE` file in this directory and are intentionally licensed separately from the BSL `LICENSE` file at the root of this repository. * Updating the license from MPL to Business Source License Going forward, this project will be licensed under the Business Source License v1.1. Please see our blog post for more details at <Blog URL>, FAQ at www.hashicorp.com/licensing-faq, and details of the license at www.hashicorp.com/bsl. * add missing license headers * Update copyright file headers to BUSL-1.1 * Update copyright file headers to BUSL-1.1 * Update copyright file headers to BUSL-1.1 * Update copyright file headers to BUSL-1.1 * Update copyright file headers to BUSL-1.1 * Update copyright file headers to BUSL-1.1 * Update copyright file headers to BUSL-1.1 * Update copyright file headers to BUSL-1.1 * Update copyright file headers to BUSL-1.1 * Update copyright file headers to BUSL-1.1 * Update copyright file headers to BUSL-1.1 * Update copyright file headers to BUSL-1.1 * Update copyright file headers to BUSL-1.1 * Update copyright file headers to BUSL-1.1 * Update copyright file headers to BUSL-1.1 --------- Co-authored-by: hashicorp-copywrite[bot] <110428419+hashicorp-copywrite[bot]@users.noreply.github.com>
2023-08-11 13:12:13 +00:00
// Copyright (c) HashiCorp, Inc.
// SPDX-License-Identifier: BUSL-1.1
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
package hcp
import (
"context"
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
"errors"
"fmt"
"net/http"
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
"net/url"
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
"reflect"
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
"regexp"
"sync"
"time"
"github.com/armon/go-metrics"
"github.com/go-openapi/runtime"
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
"github.com/hashicorp/go-hclog"
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
"github.com/hashicorp/go-retryablehttp"
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
"github.com/hashicorp/consul/agent/hcp/client"
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
"github.com/hashicorp/consul/agent/hcp/config"
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
"github.com/hashicorp/consul/agent/hcp/telemetry"
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
"github.com/hashicorp/consul/version"
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
)
var (
// internalMetricRefreshFailure is a metric to monitor refresh failures.
internalMetricRefreshFailure []string = []string{"hcp", "telemetry_config_provider", "refresh", "failure"}
// internalMetricRefreshSuccess is a metric to monitor refresh successes.
internalMetricRefreshSuccess []string = []string{"hcp", "telemetry_config_provider", "refresh", "success"}
// defaultTelemetryConfigRefreshInterval is a default fallback in case the first HCP fetch fails.
defaultTelemetryConfigRefreshInterval = 1 * time.Minute
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
)
// Ensure hcpProviderImpl implements telemetry provider interfaces.
var _ telemetry.ConfigProvider = &hcpProviderImpl{}
var _ telemetry.EndpointProvider = &hcpProviderImpl{}
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
var _ client.MetricsClientProvider = &hcpProviderImpl{}
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
// hcpProviderImpl holds telemetry configuration and settings for continuous fetch of new config from HCP.
// it updates configuration, if changes are detected.
type hcpProviderImpl struct {
// cfg holds configuration that can be dynamically updated.
cfg *dynamicConfig
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
// httpCfg holds configuration for the HTTP client
httpCfg *httpCfg
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
// Reader-writer mutexes are used as the provider is read heavy.
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
// OTEL components access telemetryConfig during metrics collection and export (read).
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
// Meanwhile, configs are only updated when there are changes (write).
rw sync.RWMutex
httpCfgRW sync.RWMutex
// running indicates if the HCP telemetry config provider has been started
running bool
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
// hcpClient is an authenticated client used to make HTTP requests to HCP.
hcpClient client.Client
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
// logger is the HCP logger for the provider
logger hclog.Logger
// testUpdateConfigCh is used by unit tests to signal when an update config has occurred
testUpdateConfigCh chan struct{}
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
}
// dynamicConfig is a set of configurable settings for metrics collection, processing and export.
// fields MUST be exported to compute hash for equals method.
type dynamicConfig struct {
disabled bool
endpoint *url.URL
labels map[string]string
filters *regexp.Regexp
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
// refreshInterval controls the interval at which configuration is fetched from HCP to refresh config.
refreshInterval time.Duration
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
}
// defaultDisabledCfg disables metric collection and contains default config values.
func defaultDisabledCfg() *dynamicConfig {
return &dynamicConfig{
labels: map[string]string{},
filters: client.DefaultMetricFilters,
refreshInterval: defaultTelemetryConfigRefreshInterval,
endpoint: nil,
disabled: true,
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
}
}
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
// httpCfg is a set of configurable settings for the HTTP client used to export metrics
type httpCfg struct {
header *http.Header
client *retryablehttp.Client
}
type HCPProviderCfg struct {
HCPClient client.Client
HCPConfig config.CloudConfigurer
}
// NewHCPProvider initializes and starts a HCP Telemetry provider.
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
func NewHCPProvider(ctx context.Context) *hcpProviderImpl {
h := &hcpProviderImpl{
// Initialize with default config values.
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
cfg: defaultDisabledCfg(),
httpCfg: &httpCfg{},
logger: hclog.FromContext(ctx),
}
return h
}
// Run starts a process that continuously checks for updates to the telemetry configuration
// by making a request to HCP. It only starts running if it's not already running.
func (h *hcpProviderImpl) Run(ctx context.Context, c *HCPProviderCfg) error {
if h.isRunning() {
return nil
}
h.rw.Lock()
h.running = true
h.rw.Unlock()
// Update the provider with the HCP configurations
h.hcpClient = c.HCPClient
err := h.updateHTTPConfig(c.HCPConfig)
if err != nil {
return fmt.Errorf("failed to initialize HCP telemetry provider: %v", err)
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
}
go h.run(ctx)
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
return nil
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
}
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
// run continuously checks for updates to the telemetry configuration by making a request to HCP.
func (h *hcpProviderImpl) run(ctx context.Context) error {
h.logger.Debug("starting telemetry config provider")
// Try to initialize config once before starting periodic fetch.
h.updateConfig(ctx)
ticker := time.NewTicker(h.cfg.refreshInterval)
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
defer ticker.Stop()
for {
select {
case <-ticker.C:
if newRefreshInterval := h.updateConfig(ctx); newRefreshInterval > 0 {
ticker.Reset(newRefreshInterval)
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
}
case <-ctx.Done():
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
return nil
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
}
}
}
// updateConfig makes a HTTP request to HCP to update metrics configuration held in the provider.
func (h *hcpProviderImpl) updateConfig(ctx context.Context) time.Duration {
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
logger := h.logger.Named("telemetry_config_provider")
if h.testUpdateConfigCh != nil {
defer func() {
select {
case h.testUpdateConfigCh <- struct{}{}:
default:
}
}()
}
if h.hcpClient == nil || reflect.ValueOf(h.hcpClient).IsNil() {
// Disable metrics if HCP client is not configured
disabledMetricsCfg := defaultDisabledCfg()
h.modifyDynamicCfg(disabledMetricsCfg)
return disabledMetricsCfg.refreshInterval
}
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
ctx, cancel := context.WithTimeout(ctx, 5*time.Second)
defer cancel()
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
logger.Trace("fetching telemetry config")
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
telemetryCfg, err := h.hcpClient.FetchTelemetryConfig(ctx)
if err != nil {
// Only disable metrics on 404 or 401 to handle the case of an unlinked cluster.
// For other errors such as 5XX ones, we continue metrics collection, as these are potentially transient server-side errors.
apiErr, ok := err.(*runtime.APIError)
if ok && apiErr.IsClientError() {
disabledMetricsCfg := defaultDisabledCfg()
h.modifyDynamicCfg(disabledMetricsCfg)
return disabledMetricsCfg.refreshInterval
}
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
logger.Error("failed to fetch telemetry config from HCP", "error", err)
metrics.IncrCounter(internalMetricRefreshFailure, 1)
return 0
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
}
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
logger.Trace("successfully fetched telemetry config")
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
// newRefreshInterval of 0 or less can cause ticker Reset() panic.
newRefreshInterval := telemetryCfg.RefreshConfig.RefreshInterval
if newRefreshInterval <= 0 {
logger.Error("invalid refresh interval duration", "refreshInterval", newRefreshInterval)
metrics.IncrCounter(internalMetricRefreshFailure, 1)
return 0
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
}
newCfg := &dynamicConfig{
filters: telemetryCfg.MetricsConfig.Filters,
endpoint: telemetryCfg.MetricsConfig.Endpoint,
labels: telemetryCfg.MetricsConfig.Labels,
refreshInterval: telemetryCfg.RefreshConfig.RefreshInterval,
disabled: telemetryCfg.MetricsConfig.Disabled,
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
}
h.modifyDynamicCfg(newCfg)
return newCfg.refreshInterval
}
// modifyDynamicCfg acquires a write lock to update new configuration and emits a success metric.
func (h *hcpProviderImpl) modifyDynamicCfg(newCfg *dynamicConfig) {
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
h.rw.Lock()
h.cfg = newCfg
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
h.rw.Unlock()
metrics.IncrCounter(internalMetricRefreshSuccess, 1)
}
// GetEndpoint acquires a read lock to return endpoint configuration for consumers.
func (h *hcpProviderImpl) GetEndpoint() *url.URL {
h.rw.RLock()
defer h.rw.RUnlock()
return h.cfg.endpoint
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
}
// GetFilters acquires a read lock to return filters configuration for consumers.
func (h *hcpProviderImpl) GetFilters() *regexp.Regexp {
h.rw.RLock()
defer h.rw.RUnlock()
return h.cfg.filters
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
}
// GetLabels acquires a read lock to return labels configuration for consumers.
func (h *hcpProviderImpl) GetLabels() map[string]string {
h.rw.RLock()
defer h.rw.RUnlock()
return h.cfg.labels
}
// IsDisabled acquires a read lock and return true if metrics are enabled.
func (h *hcpProviderImpl) IsDisabled() bool {
h.rw.RLock()
defer h.rw.RUnlock()
return h.cfg.disabled
[HCP Telemetry] Periodic Refresh for Dynamic Telemetry Configuration (#18168) * OTElExporter now uses an EndpointProvider to discover the endpoint * OTELSink uses a ConfigProvider to obtain filters and labels configuration * improve tests for otel_sink * Regex logic is moved into client for a method on the TelemetryConfig object * Create a telemetry_config_provider and update deps to use it * Fix conversion * fix import newline * Add logger to hcp client and move telemetry_config out of the client.go file * Add a telemetry_config.go to refactor client.go * Update deps * update hcp deps test * Modify telemetry_config_providers * Check for nil filters * PR review updates * Fix comments and move around pieces * Fix comments * Remove context from client struct * Moved ctx out of sink struct and fixed filters, added a test * Remove named imports, use errors.New if not fformatting * Remove HCP dependencies in telemetry package * Add success metric and move lock only to grab the t.cfgHahs * Update hash * fix nits * Create an equals method and add tests * Improve telemetry_config_provider.go tests * Add race test * Add missing godoc * Remove mock for MetricsClient * Avoid goroutine test panics * trying to kick CI lint issues by upgrading mod * imprve test code and add hasher for testing * Use structure logging for filters, fix error constants, and default to allow all regex * removed hashin and modify logic to simplify * Improve race test and fix PR feedback by removing hash equals and avoid testing the timer.Ticker logic, and instead unit test * Ran make go-mod-tidy * Use errtypes in the test * Add changelog * add safety check for exporter endpoint * remove require.Contains by using error types, fix structure logging, and fix success metric typo in exporter * Fixed race test to have changing config values * Send success metric before modifying config * Avoid the defer and move the success metric under
2023-08-01 21:20:18 +00:00
}
[CC-7042] Update and enable the HCP metrics sink in the HCP manager (#20072) * Option to set HCP client at runtime Allows us to initially set a nil HCP client for the telemetry provider and update it later. * Set telemetry provider HCP client in HCP manager Set the telemetry provider as a dependency and pass it to the manager. Update the telemetry provider's HCP client when the HCP manager starts. * Add a provider interface for the metrics client This provider will allow us to configure and reconfigure the retryable HTTP client and the headers for the metrics client. * Move HTTP retryable client to separate file Copied directly from the metrics client. * Abstract HCP specific values in HTTP client Remove HCP specific references and instead initiate with a generic TLS configuration and authentication source. * Set up HTTP client and headers in the provider Move setup from the metrics client to the HCP telemetry provider. * Update the telemetry provider in the HCP manager Initialize the provider without the HCP configs and then update it in the HCP manager to enable it. * Improve test assertion, fix method comment * Move client provider to metrics client * Stop the manager on setup error * Add separate lock for http configuration * Start telemetry provider in HCP manager * Update HCP client and config as part of Run * Remove option to set config at initialization * Simplify and clean up setting HCP configs * Add test for telemetry provider Run method * Fix race condition * Use clone of HTTP headers * Only allow initial update and run once
2024-01-16 16:46:12 +00:00
// updateHTTPConfig updates the HTTP configuration values that rely on the HCP configuration.
func (h *hcpProviderImpl) updateHTTPConfig(cfg config.CloudConfigurer) error {
h.httpCfgRW.Lock()
defer h.httpCfgRW.Unlock()
if cfg == nil {
return errors.New("must provide valid HCP configuration")
}
// Update headers
r, err := cfg.Resource()
if err != nil {
return fmt.Errorf("failed set telemetry client headers: %v", err)
}
header := make(http.Header)
header.Set("content-type", "application/x-protobuf")
header.Set("x-hcp-resource-id", r.String())
header.Set("x-channel", fmt.Sprintf("consul/%s", version.GetHumanVersion()))
h.httpCfg.header = &header
// Update HTTP client
hcpCfg, err := cfg.HCPConfig()
if err != nil {
return fmt.Errorf("failed to configure telemetry HTTP client: %v", err)
}
h.httpCfg.client = client.NewHTTPClient(
hcpCfg.APITLSConfig(),
hcpCfg,
h.logger.Named("hcp_telemetry_client"))
return nil
}
// GetHeader acquires a read lock to return the HTTP request headers needed
// to export metrics.
func (h *hcpProviderImpl) GetHeader() http.Header {
h.httpCfgRW.RLock()
defer h.httpCfgRW.RUnlock()
if h.httpCfg.header == nil {
return nil
}
return h.httpCfg.header.Clone()
}
// GetHTTPClient acquires a read lock to return the retryable HTTP client needed
// to export metrics.
func (h *hcpProviderImpl) GetHTTPClient() *retryablehttp.Client {
h.httpCfgRW.RLock()
defer h.httpCfgRW.RUnlock()
return h.httpCfg.client
}
// isRunning acquires a read lock to return whether the provider is running.
func (h *hcpProviderImpl) isRunning() bool {
h.rw.RLock()
defer h.rw.RUnlock()
return h.running
}