Skip to main content

Starting around 13:00 pacific today, we've been getting elevated levels of errors trying to verify recaptcha from our back-end.  The error we're getting back is:

{ "success": false, "error-codes": [ "timeout-or-duplicate" ]}

Here's a screenshot of stdout from our server showing instaces of "recaptcha":

As far as I can tell, we haven't changed anything that would have caused this.

The only thing I can think of is converting from a legacy to enterprise key -- that was done last Monday as a direct result of this -- https://www.googlecloudcommunity.com/gc/reCAPTCHA/ReCaptcha-Test-Key-stopped-working/m-p/860259#M346 

According to the UI of recaptcha, we're seeing elevated levels of "low risk" hits, but nothing to indicate our users are getting errors.  It's pretty much all going into error right now.

I'm not going to share our order numbers here, but the last successful order we had in Canada was 13:05 -- in that same time period, we've seen USA orders unaffected.

Please advise.

 

EDIT --  Looks like this post still isn't visible, so I can't add a reply to it.

We ended up removing recaptcha from our site completely. 

Will circle back on Monday morning to see what the status is.

This is still an issue as of 2025-01-20 10:30 -08:00

The change I had made on Friday was to stop displaying the recaptcha on the front-end completely... now I've updated this so we'll collect recaptchas, attempt to validate but log a failure if it fails.... with this change in place, I'm still seeing the timeout-or-duplicate errors.

This seems very broken... I'm very curious what changes on Friday at 13:00 pacific that started to cause all of these errors.  


Well, I figured it out... and it's completely self-inflicted.

We implemented a "threeds authentication" flow on the checkout page, and it will attempt to post back to the same endpoint once the front-end routing is completed... this will of course raise "duplicate-or-timeout"

Crisis averted!


Reply