Cron Retries Feature Launched

[Kim Lewandowski]

Hi App Engine Developers,

I'm happy to announce that we've launched a new feature for our Cron Service. We now support cron retries. You can configure your cron jobs to retry up to 5 times if they fail. Check out the documentation to learn more: Python and Java. Try it out and let me know what you think.

Kim Lewandowski | Product Manager, Google Cloud Platform | klewan...@google.com 

[Rohit Gupta]

Hey, that's very helpful. Thanks for that!

[Anastasios Hatzis]

A useful addition, thank you! May I take the opportunity and ask you something about cron-jobs? Is there a specific reason why cron-jobs spawn GET requests? Many of my task-handlers are POST methods. Maybe I miss something, but I think it would feel more natural if I could also define the desired HTTP verb in the cron-job.

[Udayan Patel]

Well, I wonder if cron provides some kind of information when it fails after max defined retry so that application or stack driver can send notification for human intervention. We have been looking into this for past couple of days but couldn't find anything we can use. On a side note, We are kind of disappointed that max retry is limited to 5 times.

Thanks

Uday

[Kim Lewandowski]

Hi Udayan,

Thanks for the feedback. For max retries, how high would you like to set this number? We can probably bump it up to 10 pretty easily, but anything over that might take us some time. For alerts, have you looked at Stackdriver's Logs-based metrics (https://cloud.google.com/logging/docs/view/logs_based_metrics)? You should be able to alert on number of times an error log is written for a given task name.