Possible regression: query which used to work all the time now returns an error

I’ve been performing the following query for a long time now:

curl --location --request POST 'https://api.github.com/graphql' \
--header 'Content-Type: application/json' \
--header 'Authorization: token <REDACTED>' \
--header 'Cookie: _gh_sess=<REDACTED>' \
--data-raw '{"query":"fragment o on Organization {  __typename  id  databaseId  login  name organizationEmail:email  websiteUrl  avatarUrl  location  updatedAt}query($login:String!,$page_size:Int!,$cursor:String){  user(login:$login){    ...on User{      id      organizations(first:$page_size,after:$cursor){        nodes{...o}        pageInfo {          endCursor          hasNextPage        }      }    }  }}","variables":{"cursor":null,"login":<REDACTED>,"page_size":25}}'

and it has always worked. Now, since yesterday and all of a sudden my workflows are broken since this query now returns the following error:

    "errors": [
            "message": "Something went wrong while executing your query. Please include `C0B7:3F3C:567BDB1:585BE87:616E7429` when reporting this issue."

I believe this to be a regression when accessing this API via an installation token. I’ve experimented with a personal access token and it works fine. As I mentioned this particular query had been working with installation tokens for a long time and no code changes were done on my side to break behaviour.

Could anyone give me some help with this? I’m pretty positive this is a bug from GitHub’s side.

Hey @sashaafm

Just wanted to let you know that I’m reviewing logs against that request ID value you provided, right now. It’s not immediately obvious what the problem is, but I’m going to raise it with the team for more assistance.

And as you called out in another thread, this topic didn’t get hasty attention and I must apologize for that. However, I do notice that you’re both a Pro member, and are part of an Enterprise level Organization. Which means you have access to both our standard Support ticketing system, and our Enterprise Support ticketing system.

While we love any level of user to participate in the forums, for something like this, it would be best if you opened up a ticket with us directly, here:

Where we have more strict SLAs for Enterprise users.

Here in the forum we encourage other users to assist with other user problems, and we aren’t easily able to identify and respond to every user’s topics individually.

Again, I’m escalating this to get further assistance and will respond when I have something more for you.

@sashaafm hmm…reviewing the logs that I have from that request_id still do seem unique from what was being discussed in:

…but your query is so similar, I’m now second guessing myself and you may be right, that it’s the same problem as those other users mentioned.

Before I start waving other teammates / dev team down, could you verify if this is ongoing? If so, how widespread/consistent is it?

If not, we could certainly chalk it up to the same experience as our friends in the aforementioned thread.

If it’s ongoing and consistent, at least one more updated request_id value would be helpful for review.