Cypress crash with "There was an error reconnecting to the Chrome DevTools protocol" #26900
Comments
|
Hi @StefanoBalzarottiNozomi Right now there doesn't seem to be enough information to reproduce the problem on our end. Unless we receive a reliable reproduction, we'll eventually have to close this issue until we can reproduce it. This does not mean that your issue is not happening - it just means that we do not have a path to move forward. Please provide a reproducible example of the issue you're encountering. Here are some tips for providing a Short, Self Contained, Correct, Example and our own Troubleshooting Cypress guide. Forking Cypress Test Tiny makes sharing a reproducible example easier to share and easier for our engineers to replicate your issues. This method also keeps the reproduction as simple as possible, which helps us eliminate potential causes and noise from the investigation. |
|
Have run into this several times since upgrading to Cypress 12.12. Once it happened 5m42s into a test run, another time 7m6s in. May not be reproducible with a minimal example. We are running cypress in a GitHub Actions runner as follows: action: startup output from cypress: |
|
We faced this too. On both local and CI (CircleCI) environments |
|
Sorry, I can't find a reproducible test case, I don't think it depends by the kind of test, because the error happens randomly with any test. Unlucky I can't share my test suite because it contains proprietary code. Maybe there is a sort of memory leak, I see the memory increase during test execution, but I am not sure if related. |
|
Hi @StefanoBalzarottiNozomi , Unfortunately without a reproducible example we will not be able to investigate this any further at this time. If you suspect memory problems may be causing this issue have you tried using the experimentalMemoryManagement flag? That may help but it is tough to say without being able to reproduce the problem. I am going to close this issue for now until such time that we have a reproducible example. If you are able to reproduce the problem and share that with us please comment back here and we can reopen the issue. |
|
We also have this lately quite often - cypress test runs became very unstable... |
|
using cypress 12.13.0, experiencing the same problem as the other posters with
downgrading it to 12.11.0 seems to have done the trick for now System specs (I highly doubt it's related to memory constraints, but here goes) |
|
@beatjost and @yarrysmod have you tried using the config flag I mentioned in my previous comment? Memory issues can happen even if the machine is robust like in your case @yarrysmod because chrome limits the amount of memory it can use. |
|
Also if either of you can provide a reproducible example demonstrating this behavior it would be extremely useful in debugging this issue. |
yes, |
|
With the However, we also have some issue as well with the dri3-extension. This is also printend every time, but seems not to affect the test result... |
|
These issues refer to two of the messages you are seeing: |
|
@yarrysmod and @beatjost , if one of you can provide a reproducible example please do comment back here and we can reopen for investigation once we have a path forward. |
|
Would it be possible to re-open this issue? Our CI is failing regularly (but not on every run) with this error. I can't provide a reproducible example, but I can say that it fails at the same point each time, when trying to start one particular test suite (which may indicate that it's triggered by something in the previous test suite, or something else, I'm not sure). |
|
@hubgit unfortunately without a reproducible example our engineers do not have enough information to be able to effectively investigate the cause of this specific issue. In order to give our engineers the best chance at recreating your problem, please create a reproducible example using a fork of Cypress Test Tiny. This gives us the best chance of seeing exactly what you are seeing and being able to investigate effectively. Are you able to provide a reproducible example? |
|
This is also happening periodically on one of our test suites. It’s hard to provide a reproducible example for what seems to be a memory leak. I think it's only happening when running tests with Electron, If this truly doesn't happen with Cypress v12.11, this PR does look fishy: https://github.com/cypress-io/cypress/pull/26573/files |
|
@Tom-Bonnike thank you for doing some detective work! That does seem like it is worth investigating further. I am going to reopen and see if we can confirm your suspicion. |
|
Hi everyone 👋, I don't think this issue is related to memory. @Tom-Bonnike, I think you are right that #26573 seems suspicious. Since we haven't been able to recreate the issue, would someone be able to run Cypress with the DEBUG logs ( |
|
Just upgraded from Cypress 10.* to the latest version (Cypress 12.14.0) and running into the same issue, both remotely in our CI pipeline and on my local development machine: |
|
Running the tests with DEBUG=cypress:server:browsers* didn't yield any useful output, so I ran it with DEBUG=cypress:* which of course printed a ton of output. I extracted the piece that pertains to the test that caused the trash this time (test 17 out of 55) and attached it here: |
|
Everytime I run it, I see the RangeError just before the crash: Invalid time value |
|
I ran the tests with Cypress versions 12.0.0, 12.8.0, 12.11.0, 12.12.0 and 12.14.0. The first three ran through while the latter two failed so it's likely the issue has been introduced in version 12.12.0. |
|
Hello, just want to confirm I also have the same issue on CircleCI, using cypress 12.14.0. Fixed when downgrading to 12.11.0 |
|
the tests either fail with the error below or the process gets stuck after all the specs are completed. The set up:
|
Please, see the log I attached to one of my the posts above. I think the invalid time value error is the culprit as it immediately precedes all the crashes I observed in the debug logs |
|
@mschile This is the debug logs from our previous run when we were using cypress v.12.12.0 I have just masked sensitive information in the logs but hoping that from this and other cypress user's logs we can investigate and implement a proper fix for this, as a workaround for now we will stick to working version 12.11.0 and this means we will miss important updates from v.12.12.0 and up. Thanks! |
|
Thanks for the debug logs! I've created a test binary, if someone who can reproduce the issue give the binary a try and see if it fixes the issue. npm install https://cdn.cypress.io/beta/npm/12.15.1/linux-x64/mschile/issue-26900_browserCriClient-9cbe5805a2344c866d27b38f5f95bec4784db058/cypress.tgz |
I tested the binary, but I am getting now the following error for each spec: Is there any specific way to use this binary ? |
|
@yajouahr, no, there isn't a specific way to use the binary. I just tried the binary again locally and was able to successfully run multiple specs. That error doesn't look like it would be related to the test binary. Do you see the same error if you use the released Cypress version 12.15.0? |
That worked for us 🎉 The issue didnt popped up in the github actions pipeline run using this test binary. The pipelines failed for versions since 12.12 (including 12.15) |
|
@WtfJoke, thanks for trying out the test binary! I've decided to change the implementation. Would you (or anyone else that can reproduce the issue) be able to give the following new test binary a try: |
|
Hi @mschile I still encountered the same error using the first test binary you provided #26900 (comment). Will try the latest test binary you provided and see if it still occurs, here's the debug logs for the same error using the first test binary btw. Thanks! |
|
I'm also getting the same error message while running Error Details Environment Details System Details Did anyone else get success in resolving this issue? |
|
@nathan5x, could you try running with the above test binary and report back with your results. |
@mschile Would you be able to publish this soon? What's the timeline for this to be available as an npm package and in Docker Hub? Thanks! |
@mschile |
With the 2nd binary the error from this issue doesn't appear. |
|
12.14.0 crashed for me for the same reasons. 12.13.0 went through, but some tests failed. However, I am excited and looking forward to the official release 12.15.1. |
|
Fyi for anyone following this issue - we released Cypress v12.16.0 today, but the fix being worked by @mschile above was not included despite the "12.15.1" version number. That is just a placeholder value and is not a guarantee of what version it might be released in. |
|
I have this issue for Cypress 12.13 after removing Cypress Dashbord --record --key from the cypress:run script |
|
if you can provide a mac os package I can test it. I can't deploy an unstable version on CI. Maybe I'll try it on a VM if you need. |
|
Confirmed working test binary |
|
Released in This comment thread has been locked. If you are still experiencing this issue after upgrading to |
Current behavior
Cypress in version 12.12 and version 12.13, randomly crash with the following error
Also reported in stackoverflow [https://stackoverflow.com/questions/76302675/cypress-there-was-an-error-reconnecting-to-the-chrome-devtools-protocol/76371091#76371091] and reproduced by different users.
Resolved downgrading to version 12.11
Desired behavior
Cypress should not crash
Test code to reproduce
Any test
Cypress Version
12.12 / 12.13
Node version
18.4.0
Operating System
MacOs 12.5 / Ubuntu 20.04.6 LTS
Debug Logs
No response
Other
No response
The text was updated successfully, but these errors were encountered: