Hi,
I think the best way here would be to read items directly from kafka,
process and store in cache and rememeber in another cache kafka stream
offset. If node crashes, your service could start from the last point
(offset).
Thanks!
-Dmitry
--
Sent from:
We have a requirement as follows:
Large streams of Time-ordered data received form over a TCP source to be
processed in a time-ordered (cannot do parallel processing of the records)
manner.
The raw data or processed data does not need to be retained/cached in the
Ignite layer.
The processed
