Genesys Cloud - Developer Community!

I am sporadically getting "Conversation Does Not Exist" as an error when I hit /api/v2/quality/evaluations/query, despite the fact that I know the conversation exists and has an evaluation. Dev says that sometimes that error actually means there is a timeout and the conversation exists. They then submitted an idea for an enhancement to fix the error message and told me that I should vote for it. 

Is there a standard that governs the accuracy of error messages in API responses? Are there similar examples where the error message doesn't match the actual result? Is there some sort of internal SLA that is creating incentive to divert issue fixes into ideas?

You didn't say what status code is being returned in the situation you described? E.g. Are you getting 404 or some other status code?

The documentation for this API lists all the response codes you can expect. And based on what you described this one sounds like the most relevant:

408 - The client did not produce a request within the server timeout limit. This can be caused by a slow network connection and/or large payloads.

i.e. If you are getting a 408 response, then I'd say it is working as expected. But if you are getting some other status code (e.g. 404) then I'd call that a bug? In the latter case I'd suggest logging a support case with Customer Care, rather than pinning your hopes on a product idea?

I am sporadically getting "Conversation Does Not Exist" as an error when I hit /api/v2/quality/evaluations/query, despite the fact that I know the conversation exists and has an evaluation. Dev says that sometimes that error actually means there is a timeout and the conversation exists. They then submitted an idea for an enhancement to fix the error message and told me that I should vote for it. 

Is there a standard that governs the accuracy of error messages in API responses? Are there similar examples where the error message doesn't match the actual result? Is there some sort of internal SLA that is creating incentive to divert issue fixes into ideas?

It's a 400 error code. This is a helpful perspective on a reasonable standard that the error code at a minimum aligns with the nature of the error. Thanks!

{"message":"Conversation does not exist in your organization","code":"quality.query.invalid.conversation","status":400,"messageParams":{},"contextId":"###","details":[],"errors":[]}

You didn't say what status code is being returned in the situation you described? E.g. Are you getting 404 or some other status code?

The documentation for this API lists all the response codes you can expect. And based on what you described this one sounds like the most relevant:

408 - The client did not produce a request within the server timeout limit. This can be caused by a slow network connection and/or large payloads.

i.e. If you are getting a 408 response, then I'd say it is working as expected. But if you are getting some other status code (e.g. 404) then I'd call that a bug? In the latter case I'd suggest logging a support case with Customer Care, rather than pinning your hopes on a product idea?

I am sporadically getting "Conversation Does Not Exist" as an error when I hit /api/v2/quality/evaluations/query, despite the fact that I know the conversation exists and has an evaluation. Dev says that sometimes that error actually means there is a timeout and the conversation exists. They then submitted an idea for an enhancement to fix the error message and told me that I should vote for it. 

Is there a standard that governs the accuracy of error messages in API responses? Are there similar examples where the error message doesn't match the actual result? Is there some sort of internal SLA that is creating incentive to divert issue fixes into ideas?