During execution, suppose there are 5 items to process in that execution.
If one of the item yields an error, it will cease operation for all these rest of the items. Should it make more sense for the rest of the items to proceed, and the 1 item that has error, to lose out?
Hey @calvintwr!
Welcome to the community 
There is an option in the Settings tab of the node called Continue on Fail. You can toggle this to true. This will execute the workflow for all the items, even if one of the items give an error.
Hi Harshil! Thanks for the quick reply! 
I see that but it allows the item that failed to continue on as well. Which would not be possible if let’s say the next node depends on the outputs of the node what is faced with some failure.
I believe commercial solutions out there like integromat will treat each item discreetly, and hence, which ever item that failed, doesn’t continue, while other items continue normally.
Seems like n8n will simply not have this function I guess? My concern is with one of those 404 or occassional 500 returns from an app, which is not uncommon and throws everything else unnecessarily off.
In that scenario you will have to set Continue on Fail to true for the other nodes as well. The other way is to use the IF node and check if there is any error.
Continue on error will for the most nodes right now continue the workflow execution but will still not process any of the remaining items of the node that did fail. The reason is, that most nodes do not implement this functionality correctly. We are currently in the process of fixing this.
You can find the currently open PR which fixes this here:
https://github.com/n8n-io/n8n/pull/1737
Hi @jan . Thanks for the info. Glad that this will be fixed and definitely I think it will be hugely beneficial.
Can I check, when you say “Continue on error will for the most nodes… continue the workflow execution… but… not process any… remaining items”, do you mean that it will continue for the item that failed, and not the rest after that?
Because I run a trial on 0.126.0, and it seems that the rest of the items will still continue.
How it behaves depends right now on the node.
For the ones where it is implemented correctly will it continue with the other items if one fails. For the ones where it is not implemented, will it stop processing after one item failed.
The linked PR above implements it correctly for the ones where it was not.
Just wanted to confirm that this fix has been deployed in v 0.131.0 and my workflow now executes as desired despite errors 
