Solved: Replication Job Invalid State Issue

cbuckley · 06-25-2024 10:32 AM

I am trying to automate some manual work and was working with the Data Fusion REST API. I was able to implement start and stop, I went to test Duplicate (get description of current job, update the name, create new replication job)

I made a mistake here and the end point for the PUT had the current replication jobs name

Now, when I view the instance for replication, I see this replication job as "Stopped". If I go into the job, the status is "Running" and if I try to stop it through the console, I get an error "Program 'program:default.<job name>.087deff9-330c-11ef-a43f-0600c721de88.worker.DeltaWorker' is not running."

I tried restarting the instance, it restarts as expected but the job is still in the same mixed state.

Any ideas what I can try? I am not able to delete the job, I can't stop the job. Not really sure what to do at this point. This is all in our dev environment.

jangemmar

Hi @cbuckley ,

Welcome to Google Cloud Community!

There seems to be a mismatch between the API call and the actual state of the replication job which causes the job unable to stop in the console.

Here is my recommendation regarding your job status mismatch in the instance and the actual job:

Identify the Duplicate Job: Since you mentioned creating a new replication job during the duplicate attempt, locate this new job.

Start the New Job: If the new job is valid, start it manually. This should activate the intended replication process.

Stop and Delete the Original Job: Once the new job is running, stop the original "Stopped" job through the console (ignoring the error message). Then, you can safely delete the original job from the Data Fusion interface.

Alternatively, you can also try to create a new Data Dusion instance and perform the replication with the correct PUT replication job’s name.

Here is the replication API reference that might help you.

View solution in original post

jangemmar