#eoghanobrien

1 messages · Page 1 of 1 (latest)

hollow novaBOT
empty yacht
#

Hi there

pseudo gazelle
#

Hi, we're seeing requests come in 1 second apart for the same tenant success page, does Stripe Checkout do some sort of retry/back-off thing after payment success with a redirect URL?

#

My understanding of the redirect process is that you are only redirected to the success URL after successful payment, even if that is an async payment. So we should only be seeing 1 incoming request from Stripe, not 3

empty yacht
#

Hmm yeah I would expect only one as well

#

Can you give me the Checkout Session ID where you see this?

pseudo gazelle
#

Yes, one minute

#

cs_live_b1tAs95MyxjtnIWfHQ7Hri1DVPfdb6vN87yrvyFcL8VxObjwubjryHDRfY

empty yacht
#

Thanks let me see what I can find from our side

#

Still looking into some things but can you clarify a bit on the above logs. This is you logging an incoming request to your success URL, correct?

pseudo gazelle
#

Correct, separate requests

#

the 2nd one seems to be the only "correct/expected" one

empty yacht
#

Okay and looks like in this case for the original redirect here the client's browser got a 404 response from you?

pseudo gazelle
#

seems to be the case

empty yacht
#

Thanks for clarifying

#

Are you seeing this consistently?

#

I'm wondering if the customer just refreshed here?

#

Overall I'm not seeing an indication of multiple redirects from our end.

hollow novaBOT
pseudo gazelle
#

it's strange that we couldn't identify a user from the session on the other 2, also the ip is different from all 3 requests

#

seeing it sporadically but sufficiently enough for it to be a more widespread issue

#

it's unlikely to be refreshed because we actually redirect to another page instantly after this

empty yacht
#

Gotcha

#

Hmmm

pseudo gazelle
#

also our pre-Stripe Checkout page, is a lengthy process where the customer designs a poster and we give a shipping estimate so 1 second per request wouldn't be enough time for completely different users to complete the process

empty yacht
#

You don't happen to have the success query params in your logs do you?

pseudo gazelle
#

Not in these logs no

empty yacht
#

Got it

pseudo gazelle
#

what were you thinking? is it possible the redirect could be without the checkout session id or something?

#

that's a scary thought

empty yacht
#

Yeah just wanted to confirm it is the same full URL each time

#

Do you have other logs of this? Are there instances where the initial redirect is a 3xx and retries still occur?

#

Do you have any retry mechanisms on your side?

uneven marsh
#

👋 also hopping in here - I'm curious, how are you identifying 'user' for each of these requests?

empty yacht
#

Like if your success URL is hit but you can't load your data for a second

#

Gotta hop away so going to leave this to Karbi for further investigation. You are in good hands!

pseudo gazelle
#

Thanks @empty yacht Hi @uneven marsh

#

User identification is based on session/auth data

uneven marsh
#

and this is only happening with some of your checkout sessions, right? not all?

pseudo gazelle
#

correct

#

i wonder if our load balancer is affecting this?

uneven marsh
#

Hmm.... that's a possibility - I was semi-thinking along those lines where something on either your end or your customer's end is intercepting the request and causing it to happen more than once

uneven marsh
#

I'm definitely only seeing us attempt to hit your success URL a single time, if that helps