Hi,
Similar to another issue I posted related to 5xx errors from Shopify, we’re also seeing an uptick recently in 400 errors coming back from Shopify for GraphQL queries and mutations we know are valid.
Using the same example as that other post (the inventorySetQuantities
mutation) we’re seeing many instances where we will send an inventory update for an inventory item successfully, then a bit later we will send that same mutation for that same inventory item (albeit with a different quantity) and receive a 400 response, then a bit later we will once again send that mutation for that inventory item and get a 200 response. Each time we get a 400 responce code we also get a response body that mentions Cloudflare (pasted below) so I’m guessing the issue lies somewhere between Shopify and Cloudflare.
Like the 5xx errors, this is another “mental blocker” in terms of fully migrating to the GraphQL APIs because it doesn’t give the team a lot of confidence moving forward. Thankfully we have retry logic built into our use of the APIs, but if the 400 errors persist longer than our retry logic then it poses problems for us. Is there any investigation or work being done to improve the reliability of these APIs prior to the deadline for forced migration?
400 response body
<html>
<head>
<title>400 Bad Request</title>
</head>
<body>
<center>
<h1>400 Bad Request</h1>
</center>
<hr>
<center>cloudflare</center>
</body>
</html>