After we upgraded to EIC, we've been encountered a problem with Import failure.
I was told when memory got pitched, the current running Instance will be restarted and It got rotated over to another instance in ready.
Meanwhile, A job running during the rotation will be disconnected and left as a success in response without a notification.
Fortunately, I was monitoring the job and was able to re-run. but we definitely would need a notification for this kind of failure.
Could I get any advice on this ?
You can create analytics report to send notification of such cases. Here job is not failed but it seems connection timeout or server got restarted for same time.