Github user matuskik commented on the pull request:
https://github.com/apache/spark/pull/4875#issuecomment-77179421
@jerryshao, I am in the process of developing my application but basically
I have a stream of events that I persist outside of Spark in Cassandra and also
fed into a window. Whenever I re-submit my application and it fails to load
from a checkpoint due to changes, I load the data from Cassandra and initialize
the window to previous state and continue streaming where I left off. The
reason is I have a large window (1-7 days) and I can't wait a day for the
window to fill up as my application will be changing fairly often. I would
welcome any other suggestions you may have.
Also, if you can think of a more efficient way to implement this, let me
know, especially in terms of garbage collecting the initial data after it goes
out of scope in the window. Do you think it would make more sense from
`WindowedDStream` to feed the initial data into the parent `DStream`'s
`generatedRDDs`? I am still wrapping my head around what exactly is executed on
the driver and worker and this is my first time writing Scala so any help is
appreciated.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
