Google App Engine 503 Error

[DOBODY GLOBAL]

Dear Google technical support

Recently on March 24, 2018, our project crashed server error code 503. We tried to fix that error but failed...  This is very important to us. Please help us.

Best regards!

[Jordan (Cloud Platform Support)]

You are seeing these errors because your requests are timing out. You can see from the picture that your requests are taking over 20 seconds to respond. If you click those errors in the Log Viewer you will most likely see the error "Request was aborted after waiting too long to attempt to service your request.". 

This error is caused by not having enough available instances to handle the incoming traffic. If your instances are too busy, new requests are forced to wait in a pending queue. If they wait too long they will be cancelled with this error as explained in the documentation. It is therefore recommended to change your scaling settings (in app.yaml or appengine-web.xml) to allow for more instances (as I assume you might be using manual scaling here where auto scaling is recommended), and to allow for concurrent requests. 

Your application code should also be optimized to respond quickly in order to allow for more requests to be handled on a smaller amount of instances. Ideally you should shard your code into different services (which use different instances), and send long running tasks to backend instances from frontend instances that respond quickly with futures to the requester (which the requester can then use to poll on the status of the long running task they started, e.g using Pub/Sub).  

- Note that Google Groups is reserved for general product discussions and not for technical support. For further technical support it is recommended to post your detailed questions to Stack Overflow using the supported Cloud tags.