Google App Engine: Request was aborted after waiting too long

[Nemanja Patrnogic]

I have a small NodeJS application backend running in a standard GAE environment which has no real world load - it only serves my personal requests when I am testing stuff.

So I was clicking through the app last night and this error message popped up:

```

Request was aborted after waiting too long to attempt to service your request.

```

These are my logs:

```

2021-03-11 06:30:06.228 CET GET /l/data 200

Api call succeeded. Returning data.

2021-03-11 06:30:52.859 CET

[start] 2021/03/11 05:30:52.858446 Quitting on terminated signal

2021-03-11 06:31:25.997 CET OPTIONS /l/orders 500

Request was aborted after waiting too long to attempt to service your request.

2021-03-11 06:31:41.132 CET OPTIONS /l/orders 500

Request was aborted after waiting too long to attempt to service your request.

2021-03-11 06:31:46.629 CET OPTIONS /l/orders 500

Request was aborted after waiting too long to attempt to service your request.

2021-03-11 06:34:44.566 CET GET /l/users 200

2021-03-11 06:34:44.736 CET

[start] 2021/03/11 05:34:44.735240 No entrypoint specified, using default entrypoint: /serve

[start] 2021/03/11 05:34:44.736120 Starting app

```

From the logs I see that my previous instance was stopped prior to the requests that triggered the error and that for the new OPTIONS requests GAE did not even attempt to start an instance. Next start was triggered almost 3 minutes later.

I am still not sure what happened here and would like to understand it because I have some production services running on GAE and this makes me a bit worried.

Thanks in advance!

[wokmou]

You might be experiencing latency from reloading instances. This can result in loading requests and additional latency as explain here[1].

You can also visit this link to know more about instance life cycle[2].

[1]https://cloud.google.com/appengine/docs/standard/java/how-instances-are-managed

[2]https://cloud.google.com/compute/docs/instances/instance-life-cycle