As far as I know debugging is hard. I was able to setup debugging in another Serverless application locally with Serverless offline. How does one go about setting up breakpoints for local clusters? Is this in the pipeline or can debugging already be done?
prestontighe
All 6 comments
I use Serverless to deploy all my functions, then I use serverless run --debug to run a local AWS API Gateway, and I can attach my Visual Studio Code to the debugger.
This FR asks for more direct debugger support: https://github.com/graphcool/framework/issues/897
kbrandwijk
on 10 Nov 2017
@kbrandwijk Ok yeah I thought about just doing it with Serverless until debugging became a thing. So all your functions are used as webhooks?
prestontighe
on 10 Nov 2017
@Toosick That depends. Functions I use in Graphcool (hooks and subscriptions) are. Most of my resolver functions are placed in my API Gateway nowadays, which I can also easily debug.
kbrandwijk
on 10 Nov 2017
@kbrandwijk how do you deploy functions with serverless in such a way that they can interface with graphcool itself? What is your workflow?
wdoug
on 13 Nov 2017
@kbrandwijk Are you aware of a write up that exists anywhere that may help someone create a setup like this? This is definitely one of my bigger pain points. Just yesterday I lost a couple hours to a silly error that would've been caught within minutes using a debugger.
fredantell
on 9 Jan 2018
debugging resolver functions is VERY time consuming.
beebase
on 11 Jan 2018
Related issues
tbrannam
路
3Comments
thomaswright
路
3Comments
dohomi
路
3Comments
notrab
路
3Comments
marktani
路
3Comments
Most helpful comment
@kbrandwijk how do you deploy functions with serverless in such a way that they can interface with graphcool itself? What is your workflow?