500 error on all POST requests to `constituent/v1/constituents`

Kevin Hathorn

Around the end of June, all of our POST requests to create constituents at `constituent/v1/constituents` suddenly started receiving 500 errors. Not a single one has succeeded since then. Other calls are fine. Queries to constituent API succeed, POSTS to the gifts api succeed. This is the only endpoint, and it's inexplicable. I took a request body from one of the failed requests, and used it to make a request via the API explorer on the dev website, and it succeeded just fine, so it does not appear it has to do with the request body (even if there was a problem with the request body, doggone sure it should not receive 500). The other calls succeeding indicate that the Auth is fine. What gives y'all?

Kevin Hathorn

Interestingly enough, if I hit the api with a request_body that triggers an error (like missing phone type), I actually get the proper 400 response with error code. But if I change the phone type back to an acceptable one, leaving all else the same, I get a 500. That makes no sense.

Anthony Gallo

@Kevin Hathorn
Anytime we are seeing 500 errors we would want to look into those as bugs. For anything bug related we would route you to our support folks to get basic troubleshooting data gathered and they would then send it to the engineering side. I will give them a heads up that this is coming so they don't just refer you back to the community.

Kevin Hathorn

@Anthony Gallo
thanks Anthony for the response and heads up. Looking forward to getting this resolved ?

Stephen Nolan

@Kevin Hathorn

We are experiencing the same bug and found that the post requests for constituents are no longer generating import ID's. Manually adding one allows you to post a constituent again but after one successful post it breaks again due to that constituent not having an import ID. Support referred me to here so hoping this gets resolved soon

Kevin Hathorn

@Anthony Gallo Any developments on this? It looks like things started to work again a few days ago, guess that means the bug was found and fixed?

Anthony Gallo

@Kevin Hathorn Did you get around to having support file a bug for this? I don't remember my team seeing anything but if that happened it is possible it was routed to another team depending on where the issue was living.

Anthony Gallo

@Kevin Hathorn it looks like support did get a bug to us and we had 2 fixes put in place so it did not just magically fix itself.