Backend instance hours display double the expected usage time

Backend
1

I’m having an issue with my App Engine backend where the instance hours counter shows twice what it should be.

I have a scheduled job that adds two identical tasks to a queue for backend processing:

// first task
taskQueue.add(withUrl("/process/workitem").param(..).header("Host", backend.getAddress("worker-backend")));

// second task  
taskQueue.add(withUrl("/process/workitem").param(..).header("Host", backend.getAddress("worker-backend")));

My queue is configured to process tasks sequentially:

<queue>
    <name>work-queue</name>
    <rate>1/m</rate>
    <max-concurrent-requests>1</max-concurrent-requests>
    <bucket-size>1</bucket-size>
    <retry-parameters>
        <task-retry-limit>1</task-retry-limit>
        <min-backoff-seconds>10</min-backoff-seconds>
        <max-backoff-seconds>200</max-backoff-seconds>
        <max-doublings>2</max-doublings>
    </retry-parameters>
</queue>

The weird thing is that when my backend runs for about 4 hours, the dashboard reports around 8 instance hours. The queue shows “Running=1” and “Tasks in Queue=2”, and I’ve confirmed my backend has only 1.0 instances configured.

Why is the billing showing double the actual runtime?

2

probly a billing bug or sumthin. app engine’s known to count partial hrs as full hrs, or even double count stuff. u should check if your backend shuts down between tasks—instances can hang around, and you’ll see charges for idle time.

3

Had the same billing issue with my App Engine backend. Google counts partial instance hours and startup overhead weird - even with one task running, your backend instance stays active between executions to avoid cold starts. Plus App Engine charges minimum billing increments, so you’ll see doubles when tasks cross billing boundaries. Check your instance logs for actual start/stop times vs what the dashboard shows. My backend was staying warm way longer than I expected, racking up charges during idle time between tasks.

4

that’s kinda strange, huh? maybe look into if your tasks are causing overlap or if there’s some hidden scaling going on. have u checked the logs for any odd behaviors?