Request was aborted after waiting too long
48 views
Skip to first unread message
Мария Кокаия
Aug 10, 2015, 8:11:59 AM8/10/15
to Google App Engine
In the period from 27 to 29 May Create Dynamic Instance did not finish the job within a specified platform time.
1). The reasons why this happens are unknown to us. Is it possible to learn more about the mechanics of creating instances?
2). How can we protect the service from such behavior in the future?
Thanks,
Mary
Nick (Cloud Platform Support)
Aug 10, 2015, 1:36:00 PM8/10/15
to Google App Engine
Hi Mary,
If you were affected by an outage or temporary issue, it's not really possible to preempt such things, and the best advice for a programmer would be to build redundancy and exception-handling and good monitoring into your app. If you believe you've been affected by an outage which violates the platform SLA, you should get in contact with Billing to determine if you're eligible for any kind of refund.
As to your specific technical issue, I'm unable to provide insight into that, and in order to get more specific insight, you'd need to open a support ticket.
If I understand your text correctly, you were using a manual scaling application and calls to create new instances using the modules service were failing. If you had any error output visible in the logs for your app at this time, that would be helpful in determining what went wrong.
At any time, if you think you've been affected by an intermittent issue on the platform, you should make an issue report to the public issue tracker as soon as possible explaining the issue, the app id, and the time-frame in which the incident occurred, so that we can potentially look into the logs for your application during the time-frame and find the cause. This is the best way to determine what went wrong short of reading your own logs in the case that it was simply a cause within your application's code.
Regards,
Nick
If you were affected by an outage or temporary issue, it's not really possible to preempt such things, and the best advice for a programmer would be to build redundancy and exception-handling and good monitoring into your app. If you believe you've been affected by an outage which violates the platform SLA, you should get in contact with Billing to determine if you're eligible for any kind of refund.
As to your specific technical issue, I'm unable to provide insight into that, and in order to get more specific insight, you'd need to open a support ticket.
If I understand your text correctly, you were using a manual scaling application and calls to create new instances using the modules service were failing. If you had any error output visible in the logs for your app at this time, that would be helpful in determining what went wrong.
At any time, if you think you've been affected by an intermittent issue on the platform, you should make an issue report to the public issue tracker as soon as possible explaining the issue, the app id, and the time-frame in which the incident occurred, so that we can potentially look into the logs for your application during the time-frame and find the cause. This is the best way to determine what went wrong short of reading your own logs in the case that it was simply a cause within your application's code.
Regards,
Nick
Dimitri Adamou
Aug 10, 2015, 7:26:06 PM8/10/15
to Google App Engine
If you are using Front-End instances, they can only last 60 seconds, you would need to create a module using Backend Instances for your heavier traffic as they can run up to 10 minutes per request and generally if you are having heavier tasks it is better to put them onto a backend instance and split them away from the main front-end traffic so that users will not feel the pinch at all.
Reply all
Reply to author
Forward
0 new messages