Assigned
Status Update
Comments
pu...@google.com <pu...@google.com>
pu...@google.com <pu...@google.com> #2
Hi , Could you please clarify the issue description or share any screen shot of the problem where you are facing issue ?
pu...@google.com <pu...@google.com> #3
Hi, its a feature request. Can you change it to feature request?
I am requesting a dataproc image version that supports spark 3.4
https://cloud.google.com/dataproc/docs/concepts/versioning/dataproc-release-2.2
I am requesting a dataproc image version that supports spark 3.4
dr...@gmail.com <dr...@gmail.com> #4
Comment has been deleted.
dd...@micron.com <dd...@micron.com> #5
Is there an update to this issue? We experience the same issue when running dataproc_v1 with pyspark batch jobs. Since the version change, we have long running jobs getting terminated after 4 hours.
da...@alltrails.com <da...@alltrails.com> #6
What worked for us was updating google-cloud-dataproc
from version 5.7.0
to 5.8.0
-- note that there were some package conflicts with other google-cloud libraries that needed to be resolved after making that version update. Also the TTL will need to be specified in units of seconds
.
Hope this helps
Description
I am attempting to specify a TTL for a Dataproc serverless PySpark Batch created via Airlfow's DataprocCreateBatchOperator using the 2.1 runtime version.
The pyspark batch config looks as follows:
When creating the batch, the operation fails with:
It looks like the
ttl
field may not be fully exposed for configuration in Dataproc serverless configs.Somewhat related issue:https://github.com/kubeflow/pipelines/issues/10190