Newrelic + Fastify: Function.prototype.apply was called on #<Object>, which is an object and not a function #2155
Comments
|
We're sorry to hear you're having trouble. A reproduction case would help us pin down what you're seeing, since there are a few moving parts (CJS-to-ESM, Fastify, Apollo). |
|
Hi @mrickard, Here is a reproduction case for both error messages: https://github.com/gevalo1/node-newrelic-issue-2155. Using fastify at all seems to cause the main error: While the logging configuration in |
|
I also added a GitHub Action in that repository to show both errors: https://github.com/gevalo1/node-newrelic-issue-2155/actions. |
|
You stated that you are seeing this problem with ESM. The provided example seems to be TypeScript. Where is the ESM reproduction? |
|
I'm not sure if it's an ESM specific problem or not, but we started seeing it when we migrated to ESM. The example I provided is indeed TypeScript, but it's also ESM. Either way I simplified the example even more by removing TypeScript entirely: https://github.com/gevalo1/node-newrelic-issue-2155. I've kept the TypeScript code under the |
|
Thank you for the clear reproduction case. It helped greatly with triaging the problem. |
|
Hi @jsumners-nr & @bizob2828, I just want to follow-up quickly regarding the reproduction, was the message expected there? |
|
This application must spawn worker threads. We do not properly function in worker threads because we do not propagate asynchronous context. In fact we don't want to make the decision of propagating context as executing in worker threads is heavily dependent on your application logic. You can opt in to loading the agent in worker threads but keep in mind that mileage may vary. |
|
The replication doesn't do anything with worker threads though? Unless either New Relic or Fastify does something with them by default? |
|
Hi, I've tried the latest version (11.16.0), and we are still getting the following warning: The replication above also throws this warning but doesn't use any worker threads. Can I safely ignore this warning and will the agent still work as expected? |
|
@gevalo1 If you take a look at the code path that logs this message, you'll see that it's logged when the agent isn't running on the main thread. As New Relic doesn't support instrumentation for applications not running on the main thread, the agent won't start. If you ignore the message, the agent won't provide instrumentation. If you override, there's no guarantee of the agent working in the same way that it works with main-thread instrumentation. |
|
Hi @mrickard, I understand that the agent doesn't support worker threads. However my question is: What is causing there to be any worker threads at all? What in my reproduction is causing this warning to be thrown? There is practically no code there and still the agent doesn't want to start? |
|
Given the code provided below (the referenced replication), I can run the application like so: $ node -r newrelic index.jsAnd get 0 lines referencing threads: $ cat newrelic_agent.log | jq -rc 'pick(.component,.module,.msg)' | rg thread
# no resultsDo you have the index.jsconst timeout = 5000;
const start = async () => {
console.log(`Starting, running for ${timeout/1000} seconds`);
await new Promise(function (resolve, _reject) {
setTimeout(function () {
console.log('Stopping, timeout expired');
resolve();
}, timeout);
});
};
start().catch((error) => {
console.log('Error', { error });
process.exit(1);
});package.json{
"dependencies": {
"newrelic": "^11.16.0"
}
}newrelic.js'use strict'
exports.config = {
app_name: ['jsumners-playground'],
license_key: 'redacted',
logging: {
level: 'trace'
},
/**
* When true, all request headers except for those listed in attributes.exclude
* will be captured for all traces, unless otherwise specified in a destination's
* attributes include/exclude lists.
*/
allow_all_headers: true,
attributes: {
/**
* Prefix of attributes to exclude from all destinations. Allows * as wildcard
* at end.
*
* NOTE: If excluding headers, they must be in camelCase form to be filtered.
*
* @name NEW_RELIC_ATTRIBUTES_EXCLUDE
*/
exclude: [
'request.headers.cookie',
'request.headers.authorization',
'request.headers.proxyAuthorization',
'request.headers.setCookie*',
'request.headers.x*',
'response.headers.cookie',
'response.headers.authorization',
'response.headers.proxyAuthorization',
'response.headers.setCookie*',
'response.headers.x*'
]
}
} |
|
I see, when I run The documentation contains a step to use the newrelic esm-loader: https://docs.newrelic.com/docs/apm/agents/nodejs-agent/installation-configuration/es-modules/#add-loader. |
|
You will definitely need that if you're using ESM. My guess is you're seeing that because the loader runs in a separate thread. I would just ignore that warning |
|
Node.js's loader API spawns threads to handle module loading. The |
|
Just to clarify still, with ignoring the warning do you mean I need to set |
Description
We're converting our project from CJS to ESM, and are encountering some issues with Newrelic & its Fastify instrumentation.
Not sure if relevant, but we also integrate @apollo/server with fastify using https://www.npmjs.com/package/@as-integrations/fastify.
Troubleshooting or NR Diag results
The above error is the main error message, but we're also seeing the following message since we migrated to ESM.
How can it be that we didn't get this message before?
Steps to Reproduce
We encountered this when migrating our project from CJS to ESM. Everything worked fine before.
If/when I'm able to create a reproduction I can share, I will link it here.
Your Environment
Node version:
20.11.1Docker Image:
node:20.11.1-alpineNewrelic
11.15.0(same error on11.14.0)Fastify:
4.26.2@fastify/cors:
9.0.1@fastify/helmet:
11.1.1@fastify/type-provider-typebox:
4.0.0@apollo/server:
4.10.2@as-integrations/fastify:
2.1.1@newrelic/apollo-server-plugin:
5.1.0newrelic.cjs
Additional context
The text was updated successfully, but these errors were encountered: