Issue with Live Migration Progress - GCP

Anita_Kis · 2025-01-13T00:30:29.971Z

Hello Team,

I am trying to initiate a live migration from a GCP VM to MongoDB Atlas, but while I am able to get past the validation step, I receive the following error after 7 minutes: “We are unable to track migration progress. Please try again later.” In the logs, I found the following entries:

Abbreviated logs:

Connection accepted - Connection accepted (IP: 52.0.74.246, port: 10823)
client metadata - Client metadata (application: MongoDB Atlas Live Migration)
Interrupted operation as its client disconnected - Operation interrupted due to client disconnection
Connection ended - Connection ended (IP: 52.0.74.246, port: 10823)
Connection accepted - Connection accepted (IP: 192.168.248.2, port: 58704)
client metadata - Client metadata (application: MongoDB Automation Agent)

The logs indicate that connections are being continuously interrupted. If you have encountered this issue before, could you please share any insights on how to resolve it?

Ankit_Luthra · 2025-01-14T19:05:04.583Z

Hi @Anita_Kis,

We’ve seen this symptom before in the context of live migrations running out of oplog space on the destination cluster. Mitigations for this specific issue include:

Allocating more oplog on the destination cluster
Disabling the embedded verification. In this case you would need to verify your migration manually. Refer: verify-data-manually
If the above steps do not help fixing the issue, please contact contact Atlas support Atlas Support