Inconsistent Sandbox Error


We are testing in sandbox environment.

Some code we run, like getting job listing with status “in_progress”, works well from a local testing environment, and it doesn’t work from our development server that has heavy testing activities. The error produced is “Invalid response received from Authorization Server. Expected JSON.”

We can’t take the go live decision with this error, especially that it is not clear what does that mean? Could you please explain it to me?


1 Like

Hello @jackygreiss,

I think I have never seen that error before. Could you please provide us with cURL examples so that we can reproduce them on our side?

Thank you,

Here is what we reached so far. We use the recommended php library/package and we found the error in it. like the attached screenshot. The library auth correctly and returns 200 sometimes and sometimes it returns response 403 (forbidden). And each requests it sends the client id and client secret to authenticate with unstable responses 200 or 403

Also, now the sandbox dashboard doesn’t login, I see a gray screen after putting the credentials. This is also weird, and I don’t know if it is related. And still our local server works well, however the development server is the one that has this issue, which is so weird.

It is confusing and we don’t know what is the reason. Is it because of the sandbox testing, or this can really occur in production environment?

ps. we don’t use caching for the token for many reasons. So we need to know why our current implementation has unstable results?

I appreciate your support.

Hello @jackygreiss,

A few comments:

  1. I just tested and the dashboard (in sandbox and prod) is loading correctly. If it doesn’t for you, can you (with Google chrome) open the “Inspector” and click on the network tab to see if any request is returning an error?
  2. I would like to see if our servers returned an error or not that could cause this “Invalid response…”. The fact you are not caching the token, should have no impact. Can you please send a direct private message with your IP address (so that we can find your logs), and also an example payload you used to create a job? With that, we’ll be able to find logs with errors.

Thank you,

Hello Sergi,

For the dashboard, it opens this morning. If the issue happens again, I will see the error like you mentioned.

And I sent an email now to see the logs, thank you.

Hello @jackygreiss,

Thank you for your patience,

We have rolled-back a change that had an impact on the rate-limiting of the requests coming from your IP. This is why the Dashboard and API were failing for you.

Could you please confirm the issue is fixed?

Thank you,

Hello Sergi,

Yes it is now fixed. Thank you.

How can we guarantee that this issue won’t happen in production? Because our client will depend on you completely for their service to be operative.

Thank you.

Hello @jackygreiss,

We do our best to keep the platform as available as possible, but like any other Tech company, from time to time we face issues that affect our clients.

Also, we try to be as reactive as possible, so if you ever encounter any issue, please let us know and we will do our best to get it addressed immediately.