After you enable Tracing Analysis for a service, Function Compute records the time consumed in the system. The time includes the cold start time, execution time of Initializer hooks, and execution time of functions. To record the time consumed for functions on the business side, you can create custom spans. For example, you can record the time taken to access services such as ApsaraDB RDS and Apsara File Storage NAS (NAS) within functions. This topic describes how to use Serverless Devs to create a custom span.
Prerequisites
Procedure
For more information about Jaeger and OpenTelemetry, see Jaeger and OpenTelemetry.
The two methods involve the following steps:
- Add Jaeger or OpenTelemetry dependencies to the function. For more information about how to add third-party dependencies, see Install third-party dependencies on Function Compute.
- Obtain the SpanContext of the trace of Function Compute from the function, create custom spans based on the SpanContext of Function Compute, and then add instrumentation before and after the code fragments whose execution time needs to be recorded.
- View the trace information on the Tracing Analysis tab of the Function Details page in the Function Compute console or in the Tracing Analysis console.
Use the SDK for Jaeger
This section describes how to use Serverless Devs to create a custom span, and create and deploy a function in a Node.js runtime environment.
- Create a code directory.
mkdir jaeger-demo
- Initialize a Node.js runtime template.
- Run the following command to go to the code directory of the project:
cd start-fc-event-nodejs12/code
- Run the following command in the code directory to initialize the package.json file:
Sample output:npm init -y
Wrote to /test/jjyy/start-fc-event-nodejs12/code/package.json: { "name": "code", "version": "1.0.0", "description": "", "main": "index.js", "scripts": { "test": "echo \"Error: no test specified\" && exit 1" }, "keywords": [], "author": "", "license": "ISC" }
- Run the following command to install Jaeger dependencies:
Sample output:npm i jaeger-client --save
added 15 packages in 1s
- Write function code.
The following code is used:
var initTracer = require('jaeger-client').initTracer; var spanContext = require('jaeger-client').SpanContext; module.exports.handler = function(event, context, callback) { console.log('invoking...',context.tracing); var config = { serviceName: 'e2e-test', reporter: { // Provide the traces endpoint; this forces the client to connect directly to the Collector and send // spans over HTTP collectorEndpoint: context.tracing.jaegerEndpoint, flushIntervalMs: 10, }, sampler: { type: "const", param: 1 }, }; var options = { tags: { 'version': 'fc-e2e-tags', }, }; var tracer = initTracer(config, options); var invokSpanContextStr = context.tracing.openTracingSpanContext; console.log('spanContext', invokSpanContextStr); var invokSpanContext = spanContext.fromString(invokSpanContextStr); var span = tracer.startSpan("CustomSpan", { childOf: invokSpanContext }); span.finish(); var params = { openTracingSpanContext: context.tracing.openTracingSpanContext, openTracingSpanBaggages: context.tracing.openTracingSpanBaggages, jaegerEndpoint: context.tracing.jaegerEndpoint } callback(null,'success') }
- Edit the s.yaml file.
- Run the following command to install dependencies:
s build --use-docker
Sample output:
[2021-11-12T18:33:43.856] [INFO ] [S-CLI] - Start ... [2021-11-12T18:33:44.677] [INFO ] [FC-BUILD] - Build artifact start... [2021-11-12T18:33:44.701] [INFO ] [FC-BUILD] - Use docker for building. [2021-11-12T18:33:44.988] [INFO ] [FC-BUILD] - Build function using image: registry.cn-beijing.aliyuncs.com/aliyunfc/runtime-nodejs12:build-1.9.21 [2021-11-12T18:33:45.011] [INFO ] [FC-BUILD] - skip pulling image registry.cn-beijing.aliyuncs.com/aliyunfc/runtime-nodejs12:build-1.9.21... [2021-11-12T18:33:47.915] [INFO ] [FC-BUILD] - Build artifact successfully. Tips for next step ====================== * Invoke Event Function: s local invoke * Invoke Http Function: s local start * Deploy Resources: s deploy End of method: build
- Run the following command to deploy the project:
Sample output:s deploy -y
[2021-11-12T18:35:26.015] [INFO ] [S-CLI] - Start ... [2021-11-12T18:35:26.633] [INFO ] [FC-DEPLOY] - Using region: cn-hangzhou [2021-11-12T18:35:26.634] [INFO ] [FC-DEPLOY] - Using access alias: default [2021-11-12T18:35:26.634] [INFO ] [FC-DEPLOY] - Using accessKeyID: LTAI4G4cwJkK4Rza6xd9**** ...... Make service fc-deploy-service success. Make function fc-deploy-service/event-nodejs12 success. [2021-11-12T18:35:37.661] [INFO ] [FC-DEPLOY] - Checking Service fc-deploy-service exists [2021-11-12T18:35:37.718] [INFO ] [FC-DEPLOY] - Checking Function event-nodejs12 exists Tips for next step ====================== * Display information of the deployed resource: s info * Display metrics: s metrics * Display logs: s logs * Invoke remote function: s invoke * Remove Service: s remove service * Remove Function: s remove function * Remove Trigger: s remove trigger * Remove CustomDomain: s remove domain fc-deploy-test: region: cn-hangzhou service: name: fc-deploy-service function: name: event-nodejs12 runtime: nodejs12 handler: index.handler memorySize: 128 timeout: 60
- Verify the results.
View the trace in the Function Compute console. The custom span that you created is connected to the spans of Function Compute.
Use OpenTelemetry
This section describes how to use Serverless Devs to create a custom span, and create and deploy a function in the Python runtime environment.
- Create a code directory.
mkdir opentelemetry-demo
- Go to the code directory.
cd opentelemeytry-demo
- Initialize a Python runtime template.
- Install OpenTelemetry dependencies.
- Run the following command to write function code:
vim index.py
The following code is used:
# -*- coding: utf-8 -*- import time from opentelemetry import trace from opentelemetry.exporter.jaeger.thrift import JaegerExporter from opentelemetry.sdk.resources import SERVICE_NAME, Resource from opentelemetry.sdk.trace import TracerProvider from opentelemetry.sdk.trace.export import SimpleSpanProcessor from opentelemetry.trace import NonRecordingSpan trace.set_tracer_provider( TracerProvider( resource=Resource.create({SERVICE_NAME: "my-helloworld-service"}) ) ) tracer = trace.get_tracer(__name__) def handler(event, context): init_tracer(context.tracing.jaeger_endpoint) span_context = get_fc_span(context.tracing.span_context) start_my_span(trace.set_span_in_context(NonRecordingSpan(span_context))) return 'hello world' def init_tracer(endpoint): # create a JaegerExporter jaeger_exporter = JaegerExporter( collector_endpoint=endpoint ) # Create a SimpleSpanProcessor and add the exporter to it span_processor = SimpleSpanProcessor(jaeger_exporter) # add to the tracer trace.get_tracer_provider().add_span_processor(span_processor) def get_fc_span(jaeger_span_context): jaeger_span_context_arr = jaeger_span_context.split(":") tid = int(jaeger_span_context_arr[0], 16) sid = int(jaeger_span_context_arr[1], 16) span_context = trace.SpanContext( trace_id=tid, span_id=sid, is_remote=True, trace_flags=trace.TraceFlags(0x01), ) return span_context def start_my_span(context): with tracer.start_as_current_span(name="fc-operation", context=context): time.sleep(0.15) with tracer.start_as_current_span("child"): time.sleep(0.1)
After you edit the function code, enter :WQ and press the Enter key to exit the editing page.
- Edit the s.yaml file.
- Run the following command to install dependencies:
s build --use-docker
Sample output:
[2021-11-12T18:53:05.818] [INFO ] [S-CLI] - Start ... [2021-11-12T18:53:06.638] [INFO ] [FC-BUILD] - Build artifact start... [2021-11-12T18:53:06.659] [INFO ] [FC-BUILD] - Use docker for building. [2021-11-12T18:53:06.888] [INFO ] [FC-BUILD] - Build function using image: registry.cn-beijing.aliyuncs.com/aliyunfc/runtime-python3.6:build-1.9.21 [2021-11-12T18:53:06.942] [INFO ] [FC-BUILD] - skip pulling image registry.cn-beijing.aliyuncs.com/aliyunfc/runtime-python3.6:build-1.9.21... [2021-11-12T18:53:10.570] [INFO ] [FC-BUILD] - Build artifact successfully. Tips for next step ====================== * Invoke Event Function: s local invoke * Invoke Http Function: s local start * Deploy Resources: s deploy End of method: build
- Run the following command to deploy the project:
Sample output:s deploy -y
[2021-11-12T18:55:02.640] [INFO ] [S-CLI] - Start ... [2021-11-12T18:55:03.455] [INFO ] [FC-DEPLOY] - Using region: cn-hangzhou [2021-11-12T18:55:03.456] [INFO ] [FC-DEPLOY] - Using access alias: default [2021-11-12T18:55:03.456] [INFO ] [FC-DEPLOY] - Using accessKeyID: LTAI4G4cwJkK4Rza6xd9**** [2021-11-12T18:55:03.457] [INFO ] [FC-DEPLOY] - Using accessKeySecret: eCc0GxSpzfq1DVspnqqd6nmYNN**** Using fc deploy type: sdk, If you want to deploy with pulumi, you can [s cli fc-default set deploy-type pulumi] to switch. [2021-11-12T18:55:04.142] [INFO ] [FC-DEPLOY] - Checking Service fc-deploy-service exists [2021-11-12T18:55:04.722] [INFO ] [FC-DEPLOY] - Checking Function event-py3 exists [2021-11-12T18:55:04.831] [INFO ] [FC-DEPLOY] - Fc detects that you have run build command for function: event-py3. [2021-11-12T18:55:04.831] [INFO ] [FC-DEPLOY] - Using codeUri: /test/jjyy/opentelemetry-demo/start-fc-event-python3/.s/build/artifacts/fc-deploy-service/event-py3 [2021-11-12T18:55:04.835] [INFO ] [FC-DEPLOY] - Fc add/append some content to your origin environment variables for finding dependencies generated by build command. { "LD_LIBRARY_PATH": "/code/.s/root/usr/local/lib:/code/.s/root/usr/lib:/code/.s/root/usr/lib/x86_64-linux-gnu:/code/.s/root/usr/lib64:/code/.s/root/lib:/code/.s/root/lib/x86_64-linux-gnu:/code/.s/root/python/lib/python2.7/site-packages:/code/.s/root/python/lib/python3.6/site-packages:/code:/code/lib:/usr/local/lib", "PATH": "/code/.s/root/usr/local/bin:/code/.s/root/usr/local/sbin:/code/.s/root/usr/bin:/code/.s/root/usr/sbin:/code/.s/root/sbin:/code/.s/root/bin:/code:/code/node_modules/.bin:/code/.s/python/bin:/code/.s/node_modules/.bin:/usr/local/bin:/usr/local/sbin:/usr/bin:/usr/sbin:/sbin:/bin", "NODE_PATH": "/code/node_modules:/usr/local/lib/node_modules", "PYTHONUSERBASE": "/code/.s/python" } Make service fc-deploy-service success. Make function fc-deploy-service/event-py3 success. [2021-11-12T18:55:10.693] [INFO ] [FC-DEPLOY] - Checking Service fc-deploy-service exists [2021-11-12T18:55:10.737] [INFO ] [FC-DEPLOY] - Checking Function event-py3 exists Tips for next step ====================== * Display information of the deployed resource: s info * Display metrics: s metrics * Display logs: s logs * Invoke remote function: s invoke * Remove Service: s remove service * Remove Function: s remove function * Remove Trigger: s remove trigger * Remove CustomDomain: s remove domain fc-deploy-test: region: cn-hangzhou service: name: fc-deploy-service function: name: event-py3 runtime: python3 handler: index.handler memorySize: 128 timeout: 60
- Verify the results.
View the trace in the Function Compute console. The custom span that you created is connected to the spans of Function Compute.