Hi All, have a doubt, What is the upgrade process for flinkdeployment without impacting the current running job?

I dont understand what you are trying to achieve. You usually change the FlinkDeployment because you want to upgrade the running application/cluster/job

No, lets say I have CRD/Operator already installed and where I deployed flinkdeployment(CR) as application mode. and the running job will process the TCP packets data. how do i upgrade or apply patch to the current running job without any impact

Any impact to what? The impact you expect is that the job will be upgraded and right? 🙂

when I upgrade the flinkdeployment using helm upgrade the JM and TM pods are creating. does it make any impact like data loss? because our application flink will receive the TCP packets continuously from the source

because our application flink will receive the TCP packets continuously from the source

If that's the case, then it look you haven't built a source with any guarantees. Because in case of a network issue, your source also wouldn't receive the TCP packets.

Our flink application cluster will receive inputs from other external source continuously and it process the same and send the output. In this case if I want to upgrade the existing flink cluster let's say, due to some bug/issue in our code. What is the update strategy do I need follow. Example: if I upgrade flink application cluster using our Helm chart JM and TM Pods recreating and getting an connection refused error at the external sources (which send to our flink cluster)

I think all Flink sources are pull based, not push based. So normally a Flink source fetches its data. You're pushing it, so I would think there isn't any guarantee that you can offer. E.g. when the Flink cluster is down, needs to upgrade etc, the data can't be pushed so it's lost.

Thanks for the explanation this helps a lot