Workflows crash n8n installation since 1.68.0

[Passana]

Bug Description

Since updating to 1.68.0, the n8n version is unstable and crashes all the time when running the workflows in the browser/testing (workflows that ran perfectly before the update)
Also they get stuck regularly with a node and do not continue (e.g. a waiting node)

To Reproduce

Until now I can not provide any information to reproduce, because I have not found a pattern yet.

I have 3 workflows that start with the same HTTP request (cloned), the first one crashes the n8n installation, so that it runs in 502 Bad Gateway after starting it, the other two run perfectly although they're all the same HTTP requests.

I tested switching to the new canvas and then returned to the old one. Maybe this caused some of this weird behaviour.

Expected behavior

Not crashing.

Operating System

Ubuntu 20.04.6 LTS

n8n Version

1.68.0

Node.js Version

20.6.1

Database

SQLite (default)

Execution mode

main (default)

[Joffcom]

Hey @Passana,

We have created an internal ticket to look into this which we will be tracking as "GHC-499"

[netroy]

can you check if the workflow settings or the n8n global settings are configured to not save executions?
I have a suspicion that this might be the same issue that we fixed in 1.69.1 earlier today.

[Passana]

It seems that only workflows where I tried to use the new canvas and then switched back to the old canva are broken.
When I start them, n8n crashes and I get 502 error for all other workflows also.

Restarting and running other workflows (old canva)works. Seems to be the new canvas that broke the workflows.
Can not use them anymore, they're broken (after switching back from new to old canva)

These are the settings of the workflow, dont think that these are responsible as the other workflows have the same and they run (old canva)

[Passana]

1.68.0 is very unstable. Workflows that ran without problems in the past now crash from time to time

[netroy]

I'm not able to reproduce this so far.

Is this happening only for manual executions, or also for production executions?

[Passana]

also for production executions. From time to time I get 502 error and n8n is down. Did not change anything than updating to 1.68.0

[netroy]

can you please check the logs or resource usage metrics to see if the server is running out of memory and crashing?

[Passana]

in the last 24hours the memory usage was not higher than 25% and CPU not higher than 20%