docs/en/setup/configuration.md - skywalking-nodejs - Git at Google

 # Configuration

 You can configure the agent with environment variables or with options passed to `agent.start()`.
 An option passed to `agent.start()` replaces the matching environment value.

 ```typescript
 import agent from 'skywalking-backend-js';

 agent.start({
   serviceName: 'checkout-service',
   serviceInstance: 'checkout-service-1',
   collectorAddress: 'oap.example.com:11800',
   secure: true,
   authorization: 'token-value',
 });
 ```

 Environment variables are read when the package is first loaded. Set them before starting the
 Node.js process. `SW_DISABLE` is checked when `start()` is called.

 ## Service and OAP connection

 | Environment variable | `agent.start()` option | Default | Description |
 | --- | --- | --- | --- |
 | `SW_AGENT_NAME` | `serviceName` | `your-nodejs-service` | Service name shown in SkyWalking. |
 | `SW_AGENT_INSTANCE` | `serviceInstance` | Host name | Service instance name shown in SkyWalking. |
 | `SW_AGENT_COLLECTOR_BACKEND_SERVICES` | `collectorAddress` | `127.0.0.1:11800` | OAP gRPC address(es). One `host:port` uses grpc-js `dns:` (all A/AAAA become endpoints; periodically re-resolved). A comma-separated list uses a static resolver with `pick_first` (literal endpoints only; no per-name DNS expansion or re-resolution). Endpoint pick order is shuffled by grpc-js (`shuffleAddressList`); the target string keeps config order so TLS authority / SNI stay on the first list entry. Under TLS, all backends must present certificates that share the needed SANs. Prefer one DNS name with multiple A/AAAA records for TLS high availability. |
 | `SW_AGENT_SECURE` | `secure` | `false` | Use TLS for the OAP gRPC connection. |
 | `SW_AGENT_AUTHENTICATION` | `authorization` | Not set | Authentication token sent to OAP. |
 | `SW_AGENT_TRACE_TIMEOUT` | `traceTimeout` | `10000` | gRPC deadline in milliseconds for trace and meter reports and service management requests. Must be a positive integer. |

 For token authentication, set the same token in OAP with `SW_AUTHENTICATION`. See
 [OAP token authentication](https://skywalking.apache.org/docs/main/next/en/setup/backend/backend-token-auth/).

 When `secure` is enabled, the agent uses the system trust store. It does not provide options for a
 custom CA, client certificate, or mutual TLS.

 ## Agent control and logging

 | Environment variable | `agent.start()` option | Default | Description |
 | --- | --- | --- | --- |
 | `SW_DISABLE` | None | Not set | Set the exact value `true` to keep the agent stopped. |
 | `SW_AGENT_LOGGING_LEVEL` | None | `warn` | Agent log level: `error`, `warn`, `info`, or `debug`. |
 | `SW_LOGGING_TARGET` | None | See below | Set to `console` to log to the console in production. |
 | `SW_AGENT_MAX_BUFFER_SIZE` | `maxBufferSize` | `1000` | Limit for active and buffered trace segments. Must be a positive integer. |
 | `SW_AGENT_DISABLE_PLUGINS` | `disablePlugins` | Empty | Comma-separated plugin file names without the `Plugin` suffix, such as `mysql,express`. |

 Outside production mode, agent logs go to the console. When `NODE_ENV=production`, logs go to
 `skywalking.log` in the working directory unless `SW_LOGGING_TARGET=console` is set.

 When the trace buffer is full, the oldest finished segment is removed. Increase the buffer only
 after checking process memory and OAP availability.

 ## Trace filtering

 | Environment variable | `agent.start()` option | Default | Description |
 | --- | --- | --- | --- |
 | `SW_IGNORE_SUFFIX` | `ignoreSuffix` | `.jpg,.jpeg,.js,.css,.png,.bmp,.gif,.ico,.mp3,.mp4,.html,.svg` | Comma-separated path suffixes that are not traced. |
 | `SW_TRACE_IGNORE_PATH` | `traceIgnorePath` | Empty | Comma-separated operation-name patterns that are not traced. Supports `?`, `*`, and `**`. |
 | `SW_HTTP_IGNORE_METHOD` | `httpIgnoreMethod` | Empty | Comma-separated HTTP methods that are not traced, such as `OPTIONS,HEAD`. |
 | `SW_COLD_ENDPOINT` | `coldEndpoint` | `false` | Add `<cold>` to the first operation name. The first span also gets a `coldStart=true` tag. |

 Ignored trace state is sent to downstream services. This stops a filtered request from creating a
 partial trace later in the call chain. See [Tracing](../features/tracing.md) for pattern examples.

 ## Database parameters

 | Environment variable | `agent.start()` option | Default | Description |
 | --- | --- | --- | --- |
 | `SW_SQL_TRACE_PARAMETERS` | `sqlTraceParameters` | `false` | Record SQL parameter values. |
 | `SW_SQL_PARAMETERS_MAX_LENGTH` | `sqlParametersMaxLength` | `512` | Maximum SQL parameter text length. |
 | `SW_MONGO_TRACE_PARAMETERS` | `mongoTraceParameters` | `false` | Record MongoDB parameter values. |
 | `SW_MONGO_PARAMETERS_MAX_LENGTH` | `mongoParametersMaxLength` | `512` | Maximum MongoDB parameter text length. |

 Parameter values can contain passwords, tokens, personal data, or other private data. Keep these
 options off unless the data is safe to collect.

 ## Runtime metrics

 | Environment variable | `agent.start()` option | Default | Description |
 | --- | --- | --- | --- |
 | `SW_AGENT_NODEJS_RUNTIME_METRICS_REPORTER_ACTIVE` | `runtimeMetricsReporterActive` | `true` | Enable Node.js runtime meters. Set to `false` to disable them. |
 | `SW_AGENT_NODEJS_RUNTIME_METRICS_REPORT_PERIOD` | `runtimeMetricsReportPeriod` | `20000` | Sample and report period in milliseconds. Must be a positive integer. |

 Old runtime metric environment names are still accepted:

 - `SW_AGENT_RUNTIME_METRICS_REPORTER_ACTIVE`
 - `SW_AGENT_NVM_METRICS_REPORTER_ACTIVE`
 - `SW_AGENT_NVM_JVM_REPORTER_ACTIVE`
 - `SW_AGENT_RUNTIME_METRICS_REPORT_PERIOD`
 - `SW_AGENT_NVM_METRICS_REPORT_PERIOD`
 - `SW_AGENT_NVM_JVM_METRICS_REPORT_PERIOD`

 The old program options `nvmMetricsReporterActive`, `nvmJvmReporterActive`,
 `nvmMetricsReportPeriod`, and `nvmJvmMetricsReportPeriod` are also accepted. Use the current names
 for new deployments.

 ## AWS options

 | Environment variable | `agent.start()` option | Default | Description |
 | --- | --- | --- | --- |
 | `SW_AWS_LAMBDA_FLUSH` | `awsLambdaFlush` | `2` | Seconds between Lambda calls after which the wrapper flushes on exit. `0` means always; `-1` means never. |
 | `SW_AWS_LAMBDA_CHAIN` | `awsLambdaChain` | `false` | Add trace context to an AWS Lambda invoke payload. Use only when the caller and called Lambda are instrumented. |
 | `SW_AWS_SQS_CHECK_BODY` | `awsSQSCheckBody` | `false` | Also look in an incoming SQS message body for trace context passed through SNS. |

 Read [AWS SDK v2](../plugins/aws-sdk-v2.md) and [Serverless](../advanced/serverless.md) before
 enabling these options.
	# Configuration

	You can configure the agent with environment variables or with options passed to `agent.start()`.
	An option passed to `agent.start()` replaces the matching environment value.

	```typescript
	import agent from 'skywalking-backend-js';

	agent.start({
	serviceName: 'checkout-service',
	serviceInstance: 'checkout-service-1',
	collectorAddress: 'oap.example.com:11800',
	secure: true,
	authorization: 'token-value',
	});
	```

	Environment variables are read when the package is first loaded. Set them before starting the
	Node.js process. `SW_DISABLE` is checked when `start()` is called.

	## Service and OAP connection

	\| Environment variable \| `agent.start()` option \| Default \| Description \|
	\| --- \| --- \| --- \| --- \|
	\| `SW_AGENT_NAME` \| `serviceName` \| `your-nodejs-service` \| Service name shown in SkyWalking. \|
	\| `SW_AGENT_INSTANCE` \| `serviceInstance` \| Host name \| Service instance name shown in SkyWalking. \|
	\| `SW_AGENT_COLLECTOR_BACKEND_SERVICES` \| `collectorAddress` \| `127.0.0.1:11800` \| OAP gRPC address(es). One `host:port` uses grpc-js `dns:` (all A/AAAA become endpoints; periodically re-resolved). A comma-separated list uses a static resolver with `pick_first` (literal endpoints only; no per-name DNS expansion or re-resolution). Endpoint pick order is shuffled by grpc-js (`shuffleAddressList`); the target string keeps config order so TLS authority / SNI stay on the first list entry. Under TLS, all backends must present certificates that share the needed SANs. Prefer one DNS name with multiple A/AAAA records for TLS high availability. \|
	\| `SW_AGENT_SECURE` \| `secure` \| `false` \| Use TLS for the OAP gRPC connection. \|
	\| `SW_AGENT_AUTHENTICATION` \| `authorization` \| Not set \| Authentication token sent to OAP. \|
	\| `SW_AGENT_TRACE_TIMEOUT` \| `traceTimeout` \| `10000` \| gRPC deadline in milliseconds for trace and meter reports and service management requests. Must be a positive integer. \|

	For token authentication, set the same token in OAP with `SW_AUTHENTICATION`. See
	[OAP token authentication](https://skywalking.apache.org/docs/main/next/en/setup/backend/backend-token-auth/).

	When `secure` is enabled, the agent uses the system trust store. It does not provide options for a
	custom CA, client certificate, or mutual TLS.

	## Agent control and logging

	\| Environment variable \| `agent.start()` option \| Default \| Description \|
	\| --- \| --- \| --- \| --- \|
	\| `SW_DISABLE` \| None \| Not set \| Set the exact value `true` to keep the agent stopped. \|
	\| `SW_AGENT_LOGGING_LEVEL` \| None \| `warn` \| Agent log level: `error`, `warn`, `info`, or `debug`. \|
	\| `SW_LOGGING_TARGET` \| None \| See below \| Set to `console` to log to the console in production. \|
	\| `SW_AGENT_MAX_BUFFER_SIZE` \| `maxBufferSize` \| `1000` \| Limit for active and buffered trace segments. Must be a positive integer. \|
	\| `SW_AGENT_DISABLE_PLUGINS` \| `disablePlugins` \| Empty \| Comma-separated plugin file names without the `Plugin` suffix, such as `mysql,express`. \|

	Outside production mode, agent logs go to the console. When `NODE_ENV=production`, logs go to
	`skywalking.log` in the working directory unless `SW_LOGGING_TARGET=console` is set.

	When the trace buffer is full, the oldest finished segment is removed. Increase the buffer only
	after checking process memory and OAP availability.

	## Trace filtering

	\| Environment variable \| `agent.start()` option \| Default \| Description \|
	\| --- \| --- \| --- \| --- \|
	\| `SW_IGNORE_SUFFIX` \| `ignoreSuffix` \| `.jpg,.jpeg,.js,.css,.png,.bmp,.gif,.ico,.mp3,.mp4,.html,.svg` \| Comma-separated path suffixes that are not traced. \|
	\| `SW_TRACE_IGNORE_PATH` \| `traceIgnorePath` \| Empty \| Comma-separated operation-name patterns that are not traced. Supports `?`, ``, and `*`. \|
	\| `SW_HTTP_IGNORE_METHOD` \| `httpIgnoreMethod` \| Empty \| Comma-separated HTTP methods that are not traced, such as `OPTIONS,HEAD`. \|
	\| `SW_COLD_ENDPOINT` \| `coldEndpoint` \| `false` \| Add `<cold>` to the first operation name. The first span also gets a `coldStart=true` tag. \|

	Ignored trace state is sent to downstream services. This stops a filtered request from creating a
	partial trace later in the call chain. See [Tracing](../features/tracing.md) for pattern examples.

	## Database parameters

	\| Environment variable \| `agent.start()` option \| Default \| Description \|
	\| --- \| --- \| --- \| --- \|
	\| `SW_SQL_TRACE_PARAMETERS` \| `sqlTraceParameters` \| `false` \| Record SQL parameter values. \|
	\| `SW_SQL_PARAMETERS_MAX_LENGTH` \| `sqlParametersMaxLength` \| `512` \| Maximum SQL parameter text length. \|
	\| `SW_MONGO_TRACE_PARAMETERS` \| `mongoTraceParameters` \| `false` \| Record MongoDB parameter values. \|
	\| `SW_MONGO_PARAMETERS_MAX_LENGTH` \| `mongoParametersMaxLength` \| `512` \| Maximum MongoDB parameter text length. \|

	Parameter values can contain passwords, tokens, personal data, or other private data. Keep these
	options off unless the data is safe to collect.

	## Runtime metrics

	\| Environment variable \| `agent.start()` option \| Default \| Description \|
	\| --- \| --- \| --- \| --- \|
	\| `SW_AGENT_NODEJS_RUNTIME_METRICS_REPORTER_ACTIVE` \| `runtimeMetricsReporterActive` \| `true` \| Enable Node.js runtime meters. Set to `false` to disable them. \|
	\| `SW_AGENT_NODEJS_RUNTIME_METRICS_REPORT_PERIOD` \| `runtimeMetricsReportPeriod` \| `20000` \| Sample and report period in milliseconds. Must be a positive integer. \|

	Old runtime metric environment names are still accepted:

	- `SW_AGENT_RUNTIME_METRICS_REPORTER_ACTIVE`
	- `SW_AGENT_NVM_METRICS_REPORTER_ACTIVE`
	- `SW_AGENT_NVM_JVM_REPORTER_ACTIVE`
	- `SW_AGENT_RUNTIME_METRICS_REPORT_PERIOD`
	- `SW_AGENT_NVM_METRICS_REPORT_PERIOD`
	- `SW_AGENT_NVM_JVM_METRICS_REPORT_PERIOD`

	The old program options `nvmMetricsReporterActive`, `nvmJvmReporterActive`,
	`nvmMetricsReportPeriod`, and `nvmJvmMetricsReportPeriod` are also accepted. Use the current names
	for new deployments.

	## AWS options

	\| Environment variable \| `agent.start()` option \| Default \| Description \|
	\| --- \| --- \| --- \| --- \|
	\| `SW_AWS_LAMBDA_FLUSH` \| `awsLambdaFlush` \| `2` \| Seconds between Lambda calls after which the wrapper flushes on exit. `0` means always; `-1` means never. \|
	\| `SW_AWS_LAMBDA_CHAIN` \| `awsLambdaChain` \| `false` \| Add trace context to an AWS Lambda invoke payload. Use only when the caller and called Lambda are instrumented. \|
	\| `SW_AWS_SQS_CHECK_BODY` \| `awsSQSCheckBody` \| `false` \| Also look in an incoming SQS message body for trace context passed through SNS. \|

	Read [AWS SDK v2](../plugins/aws-sdk-v2.md) and [Serverless](../advanced/serverless.md) before
	enabling these options.