Class RetryConfig (2.6.4)

RetryConfig(mapping=None, *, ignore_unknown_fields=False, **kwargs)

Settings that determine the retry behavior.

By default, if a job does not complete successfully (meaning that an acknowledgement is not received from the handler, then it will be retried with exponential backoff according to the settings in RetryConfig.


retry_count int
The number of attempts that the system will make to run a job using the exponential backoff procedure described by max_doublings. The default value of retry_count is zero. If retry_count is zero, a job attempt will *not* be retried if it fails. Instead the Cloud Scheduler system will wait for the next scheduled execution time. If retry_count is set to a non-zero number then Cloud Scheduler will retry failed attempts, using exponential backoff, retry_count times, or until the next scheduled execution time, whichever comes first. Values greater than 5 and negative values are not allowed.
max_retry_duration google.protobuf.duration_pb2.Duration
The time limit for retrying a failed job, measured from time when an execution was first attempted. If specified with retry_count, the job will be retried until both limits are reached. The default value for max_retry_duration is zero, which means retry duration is unlimited.
min_backoff_duration google.protobuf.duration_pb2.Duration
The minimum amount of time to wait before retrying a job after it fails. The default value of this field is 5 seconds.
max_backoff_duration google.protobuf.duration_pb2.Duration
The maximum amount of time to wait before retrying a job after it fails. The default value of this field is 1 hour.
max_doublings int
The time between retries will double max_doublings times. A job's retry interval starts at min_backoff_duration, then doubles max_doublings times, then increases linearly, and finally retries retries at intervals of max_backoff_duration up to retry_count times. For example, if min_backoff_duration is 10s, max_backoff_duration is 300s, and max_doublings is 3, then the a job will first be retried in 10s. The retry interval will double three times, and then increase linearly by 2^3 \* 10s. Finally, the job will retry at intervals of max_backoff_duration until the job has been attempted retry_count times. Thus, the requests will retry at 10s, 20s, 40s, 80s, 160s, 240s, 300s, 300s, .... The default value of this field is 5.