Navigation

Live Migration Troubleshooting

Before the Live Migration process begins, Atlas performs a validation check to ensure that all the necessary form fields and parameters are functional and correct. If any parameters are invalid, Atlas returns an error and Live Migration does not proceed.

Listed below are some common Live Migration validation errors and suggestions for what to check if you encounter them.

Common Live Migration Validation Errors

Error Remediation
Could not reach specified source
  1. Ensure that the correct subnet ranges are whitelisted on the source cluster. The four required subnet ranges can be found in the Live Migration modal window.

  2. Confirm that the hostname provided is resolvable to a public IP address. At a command prompt, use one of the following commands:

    nslookup <hostname>
    ping <hostname>
    
  3. Ensure that you are not using a VPC Peering Connection, which is not compatible with Live Migration. If a VPC Peering Connection is your only option, use mongomirror instead.

Could not resolve hostname No IP address was found for the given hostname. Confirm that the given hostname is correct and publicly accessible.
Invalid SSL options provided

If you are using SSL:

  • Confirm that your SSL certificate is complete and correctly copied to the Live Migration modal window.
  • Confirm that the Is SSL enabled? toggle switch is in the Yes position.

If you are not using SSL:

  • Check your connection string and confirm that the ssl query parameter is not present. If --ssl is part of your connection string, your cluster requires an SSL connection.
  • Confirm that the Is SSL enabled? toggle switch is in the No position.
The username or password is not correct

Confirm your credentials in the mongo shell with the following commands:

use admin
db.getUser("<username>");

If the issue persists, update the MongoDB user’s password.

User not authorized to execute command In order to perform the Live Migration procedure, the MongoDB user must have sufficient system privileges. See Source Cluster Security for details.
Source disk usage is too large for destination Different Atlas service tiers have different amounts of disk space available. Ensure that your Atlas cluster has enough disk space for all the data on your source cluster. See Create a New Cluster for details on cluster sizing.
Source appears to be a standalone Your source deployment must be a MongoDB replica set. If your source deployment is currently a standalone node, convert it to a single-node replica set before performing Live Migration.
Unable to process the provided CA file Confirm that your CA file is complete and correctly pasted into the Live Migration modal window.

Common Post-Validation Errors

Error Remediation
Could not retrieve the latest oplog entry from the source: not found
  • If the source is a replica set, make sure you have read access on the local database.
  • If the source is a standalone instance, convert it to a replica set before proceeding with the migration.
  • Confirm that the source cluster has a readable oplog. On some hosted services, such as Compose.io, the oplog is a feature which must be enabled.
  • If you still cannot access the oplog, use mongorestore instead to import your data into Atlas.
Could not determine if –host is a replica set: error connecting to db server: no reachable servers
  • Ensure that every IP address the Live Migration service requires is whitelisted for your source cluster.
  • Confirm that the IP address or DNS hostname you provided resolves to an IP address that is publicly accessible.
Error applying oplog entries during initial sync: renameCollection command encountered during initial sync. Please restart mongomirror.

Renaming a collection on the source cluster during Live migration may trigger this error.

  • Ensure that no users or applications rename any collections while the Live Migration is taking place.
  • Aggregation operations which use $out may trigger this error. Ensure that no $out operations occur during the Live Migration procedure.
Unsupported index error Certain types and configurations of indexes which were allowable in earlier versions of MongoDB are no longer supported in more recent versions. Check the release notes for the MongoDB version on your destination cluster for possible conflicts. If necessary, drop any indexes which cause errors and rebuild them after migration is complete.