Configuring Traces for Java SDK
As you use AWS Lambda functions to deploy serverless, the entire AWS SDK becomes complementary to your application architecture. That means when developing and maintaining your application, it’s necessary to monitor how your Lambda functions are interacting with external AWS SDK services, such as DynamoDB, SQS, SNS, Kinesis, Firehose, S3, Lambda, Athena, etc. Thundra’s Java agent manages to integrate with AWS SDK, providing in-depth monitoring capabilities through trace spans. This functionality is enabled by default, but can also be disabled through configuration if needed.
You can mask SQS messages sent on the client’s side, which calls AWS SDK by setting the following environment variable:
thundra_agent_trace_integrations_aws_sqs_message_mask: true
You can mask SNS messages sent on the client’s side, which calls AWS SDK by setting the following environment variable:
thundra_agent_trace_integrations_aws_sns_message_mask: true
You can mask DynamoDB statements on the client’s side, which calls AWS SDK by setting the following environment variable:
thundra_agent_trace_integrations_aws_dynamodb_statement_mask: true
You can mask Lambda invocation payloads on the client’s side, which calls AWS SDK by setting the following environment variable:
thundra_agent_trace_integrations_aws_lambda_payload_mask: true
You can mask a request body on the client’s side by setting the following environment variable:
thundra_agent_trace_integrations_http_body_mask: true
You can mask Athena statements on the client’s side, which calls AWS SDK by setting the following environment variable:
thundra_agent_trace_integrations_aws_athena_statement_mask: true
By default, Kinesis records on the client’s side are not traced. You can enable tracing Kinesis records by setting the following environment variable:
thundra_agent_trace_integrations_aws_kinesis_record_unmask: true
By default, Firehose records on the client’s side are not traced. You can enable tracing Firehose records by setting the following environment variable:
thundra_agent_trace_integrations_aws_firehose_record_unmask: true
Tracing AWS SDK calls is enabled by default, but can also be disabled through configuration if needed. You just need to set
thundra_agent_trace_integration_aws_disable to true.As fast access to data is a crucial attribute for several applications, you may use Redis clients (such as Redisson) with your Java Lambda functions in your application architectures. However, monitoring how your application interacts with Redis is quite the task, especially when clusters and nodes come into play. Thundra's Java agent integrates with various Java Redis clients to provide in-depth monitoring. Currently, the following Redis clients are supported:
- Jedis
- Lettuce
- Redisson
You can mask Redis calls by setting the following environment variable:
thundra_agent_trace_integrations_redis_command_mask: true
Tracing Redis calls is enabled by default, but can also be disabled through configuration if needed. You can disable your integrations via the Lambda environment variable. You just need to set
thundra_agent_trace_integration_redis_disable to true.Monitoring HTTP requests is an important part of understanding how your system operates. Depending on how you implement your application, you may make several API calls; thus, you would need to be able to see how these interactions with your API are executed. Thundra's Java agent integrates with HTTP to trace HTTP requests over HTTP clients. Currently, the following HTTP clients are supported:
- Apache HTTP Client
You can mask a HTTP request body on the client’s side by setting the following environment variable:
thundra_agent_trace_integrations_http_body_mask: true
Thundra traces HTTP calls by default. However, you may disable it using environment variables. You can set the
thundra_agent_trace_integrations_http_disable environment variable to true for this purpose.Data storage is an integral part of any application, and so monitoring how your Lambda functions in this context is of paramount importance. Thundra's Java agent traces query executions over JDBC
java.sql.Statements and java.sql.PreparedStatement. Thundra is able to monitor JDBC responses for various database engines, which include the following:- PostgreSQL
- MySQL
- MariaDB
- Microsoft SQL Server
You can mask JDBC queries on the client’s side by setting the following environment variable:
thundra_agent_trace_integrations_rdb_statement_mask: true
Tracing JDBC queries is enabled by default, but can also be disabled if needed by setting the
thundra_agent_trace_integrations_rdb_disable environment variable to true.Thundra’s Java agent integrates with various Java Elasticsearch clients to provide in-depth monitoring. Currently, the following Elasticsearch clients are supported:
- High level Rest client (6.x versions)
You can mask an Elasticsearch body on the client’s side by setting the following environment variable:
thundra_agent_trace_integrations_elasticsearch_body_mask: true
Tracing Elasticsearch responses is enabled by default, but can also be disabled through configuration if needed. You can disable your integrations with the Lambda environment variable. You just need to set
thundra_agent_trace_integrations_elasticsearch_disable to true.Thundra’s Java agent integrates with various Java MongoDB clients to provide in-depth monitoring. Currently, the following MongoDB clients are supported:
- MongoDB Java Driver (3.1+ versions)
You can mask MongoDB commands on the client’s side by setting the following environment variable:
thundra_agent_trace_integrations_mongodb_command_mask: true
Tracing MongoDB operations is enabled by default, but can also be disabled through configuration if needed. You can disable your integrations with the Lambda environment variable. You just need to set
thundra_agent_trace_integrations_mongodb_disable to true.By default, requests and responses are traced, but can be disabled with environment variable configuration.
To disable tracing requests, set the
thundra_agent_lambda_trace_request_skip environment variable to true.Configuration via Environment Variable
thundra_agent_lambda_trace_request_skip: true
To disable tracing responses, set the thundra_agent_lambda_trace_response_skip environment variable to true.
Configuration via Environment Variable
thundra_agent_lambda_trace_response_skip: true
By default, incoming Kinesis records for a triggered Lambda are not traced. You can enable sending Kinesis records in a request by setting the following environment variable:
thundra_agent_lambda_trace_kinesis_request_enable: true
By default, incoming Firehose records for a triggered Lambda are not traced. You can enable sending Firehose records in a request by setting the following environment variable:
thundra_agent_lambda_trace_kinesis_request_enable: true
By default, incoming CloudWatch logs for a triggered Lambda are not traced. You can enable sending CloudWatch logs in a request by setting the following environment variable:
thundra_agent_lambda_trace_cloudwatchlog_request_enable: true
With request tracing customization support, you can specify which fields of the request should be traced or how the request should be traced.
For example, if you only want to trace the
id field of the UserGetRequest, you can customize the request masking behavior of trace support by overriding the trace plugin, as shown in the following example:UserGetHandler
....
import com.amazonaws.services.lambda.runtime.Context;
import io.thundra.agent.lambda.core.handler.request.LambdaRequestHandler;
import io.thundra.agent.lambda.core.handler.LambdaHandlerConfig;
import io.thundra.agent.lambda.core.handler.plugin.LambdaHandlerPlugin;
import io.thundra.agent.lambda.trace.handler.plugin.TraceLambdaHandlerPlugin;
...
public class UserGetHandler
implements LambdaRequestHandler<UserGetRequest, UserGetResponse> {
...
@Override
public LambdaHandlerConfig<UserGetRequest, UserGetResponse> getConfig() {
return new LambdaHandlerConfig<UserGetRequest, UserGetResponse>() {
@Override
public List<LambdaHandlerPlugin<UserGetRequest, UserGetResponse>> getPlugins() {
return Arrays.asList(
new TraceLambdaHandlerPlugin<UserGetRequest, UserGetResponse>() {
...
@Override
protected Object maskRequest(UserGetRequest request) {
return new HashMap<String, Object>() {{
put("id", request.getId());
}};
}
...
});
}
};
}
...
}
Note: This is also supported over
LambdaRequestStreamHandler.With response tracing customization support, you can specify which fields of the response should be traced or how the response should be traced.
For example, if you only want to trace the
id field of the User in the UserGetResponse, you can customize response masking behavior of trace support by overriding the trace plugin, as shown in the following example:UserGetHandler
...
import io.thundra.agent.lambda.core.handler.request.LambdaRequestHandler;
import io.thundra.agent.lambda.core.handler.LambdaHandlerConfig;
import io.thundra.agent.lambda.core.handler.plugin.LambdaHandlerPlugin;
import io.thundra.agent.lambda.trace.handler.plugin.TraceLambdaHandlerPlugin;
...
public class UserGetHandler
implements LambdaRequestHandler<UserGetRequest, UserGetResponse> {
...
@Override
public LambdaHandlerConfig<UserGetRequest, UserGetResponse> getConfig() {
return new LambdaHandlerConfig<UserGetRequest, UserGetResponse>() {
@Override
public List<LambdaHandlerPlugin<UserGetRequest, UserGetResponse>> getPlugins() {
return Arrays.asList(
new TraceLambdaHandlerPlugin<UserGetRequest, UserGetResponse>() {
...
@Override
protected Object maskResponse(UserGetResponse response) {
return new HashMap<String, Object>() {{
put("id", response.getUser() != null ? response.getUser().getId() : null);
}};
}
...
});
}
};
}
...
}
Note: This is also supported over
LambdaRequestStreamHandler.Last modified 2yr ago