7.x causes invalid range error (process.uptime)

[Venipa]

Is there an existing issue for this?

• I have checked for existing issues https://github.com/getsentry/sentry-javascript/issues
• I have reviewed the documentation https://docs.sentry.io/
• I am using the latest SDK release https://github.com/getsentry/sentry-javascript/releases

How do you use Sentry?

Sentry Saas (sentry.io)

Which package are you using?

@sentry/node

SDK Version

7.14.0

Framework Version

@sentry/node@7.14.0

Link to Sentry event

https://sentry.io/organizations/koellisch-gmbh/issues/3591474722/events/c8b2672f4f444952a11d7e012d9302bd/

RangeError: Invalid time value
  ?, in Date.toISOString
  File "C:\home\site\wwwroot\node_modules\@sentry\node\cjs\integrations\context.js", line 177, col 64, in getDeviceContext
  File "C:\home\site\wwwroot\node_modules\@sentry\node\cjs\integrations\context.js", line 95, col 25, in Context._getContexts
  File "C:\home\site\wwwroot\node_modules\@sentry\node\cjs\integrations\context.js", line 51, col 46, in Context.addContext

Steps to Reproduce

0. prerequisites: node v16.15.1, azure function js runtime 4.x (supports node 14.x - 16.x)
1. setup Azure functions (with typescript)
2. cause error/exception

Expected Result

exception to be submitted

Actual Result

event id is being generated but the exception is not being submitted

might be related to the pr from #5512

[Venipa]

temporary fix is to rollback to v6 (6.19.7)

[Lms24]

Hi @Venipa, thanks for writing in and reporting!

I took a quick look at the event and I think you're right with #5512 causing this. Might be related to this line:

sentry-javascript/packages/node/src/integrations/context.ts

Line 195 in 7acff3a

device.boot_time = new Date(Date.now() - os.uptime() * 1000).toISOString();

Perhaps os.uptime() is undefined or returns undefined in azure functions? I tried it out in Node and got exactly the same error when replacing os.uptime with undefined.

I'll open up a quick bugfix PR to get you back to v7.

[Lms24]

I wanted to get to the bottom of this a little more because of @timfish's comment in the bugfix PR.

I therefore created an Azure account and made a little test Http trigger function (Node 16, Windows, Azure Runtime 4). Turns out, os.uptime seems to be the only function we access from os that returns undefined (the function itself is defined). All other os functions we need return valid values.

Excerpt from the execution logs:

2022-09-30T12:40:06Z   [Information]   uptime: undefined
2022-09-30T12:40:06Z   [Information]   arch: ia32
2022-09-30T12:40:06Z   [Information]   totalmem: 3220754432
2022-09-30T12:40:06Z   [Information]   freemem: 1614860288
2022-09-30T12:40:06Z   [Information]   cpus: [
  {
    model: 'Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz',
    speed: 2594,
    times: {
      user: 4724156,
      nice: 0,
      sys: 2849796,
      idle: 321804453,
      irq: 62203
    }
  },
  {
    model: 'Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz',
    speed: 2594,
    times: {
      user: 4391875,
      nice: 0,
      sys: 2726140,
      idle: 322260187,
      irq: 32218
    }
  }
]

[Lms24]

Since it came up: process.uptime() returns a valid number afaict:

2022-09-30T12:51:38Z   [Information]   process.uptime: 30.3910407

[jlev]

This is causing a lot of false positive events in my Sentry account. I've muted the issue, but I'm worried about blowing through my quota this month. Any timeline on getting a new version of the node package out? Didn't see this change referenced in 7.16.0

[Venipa]

temporary fix is to rollback to v6 (6.19.7)

@jlev

[timfish]

This fix made it into 7.14.1:
https://github.com/getsentry/sentry-javascript/blob/master/CHANGELOG.md#7141

• fix(node): Only set DeviceContext.boot_time if os.uptime() is valid (fix(node): Only set DeviceContext.boot_time if os.uptime() is valid #5859)

If you are getting issues with any version after this, please open a new issue!

[Lms24]

Hi @jlev the fix to this (#5226) was released in 7.14.1. Is this still a problem in 7.16.0?

[jlev]

Aha! Missed that point release. Thanks for the quick response, everyone.

[joelcollyer]

We just started to run into a very similar error with @sentry/node v7.49.0 and @sentry/tracing v7.49.0 in our Express API.

Without any code changes, Sentry suddenly stopped logging data (around Apr 14, 2023), and this issue started to appear in all environments:

Based on a recommendation from Sentry Support, we added new Sentry.Integrations.Context({ device: false }), to our Sentry.init and data started to correctly appear again.

Our whole Sentry.init now looks like this (at time of writing):

Sentry.init({
  environment,
  dsn: process.env.SENTRY_DSN,
  integrations: [
    new Sentry.Integrations.Context({ device: false }),
    new Sentry.Integrations.Http({ tracing: true }),
    new Tracing.Integrations.Express({ app })
  ],
  tracesSampleRate: 1.0
});
app.use(Sentry.Handlers.requestHandler());
app.use(Sentry.Handlers.tracingHandler());

[Lms24]

Hi @joelcollyer, would you mind opening a new issue? I'm not sure how related this is to the original issue, given that you're experiencing this in Express and the original issue was in Azur functions. Thanks :)