Can I Take a VM Snapshot Of My N-able N-central Server?

Description

Backing up your N-able N-central server is critical to ensuring good disaster recovery. Virtual Machines have tools built in that can provide increased flexibility and power of backup technology, but how does it interact with N-able N-central, and what are best practices for VM Snapshots?

Solution

Disaster recovery is incredibly important and our very first recommendation is to ensure that the N-able N-central backups are running daily, and preferably that the backups are being sent off-server by the built-in FTP process if possible. Keeping a recent copy of the backup can help with disaster recovery should a critical hardware failure make the backups non-recoverable. Our recommended path when all else fails is to reinstall N-able N-central and restore using our backups, however, VM Snapshots provide power and versatility, as well as fast recovery times, that are difficult for other processes to compete with.

When using the built-in snapshot features of VMware and Hpyer-V there are a few things that have a chance to cause consistency problems or corruption with N-able N-central due to the high-volume database processes involved. In simplest terms, these snapshot processes will perform a freeze action on the Guest (in this case N-able N-central) to copy the snapshot information while maintaining a delta of the ongoing processes, then after completing the copy it will consolidate the delta with the frozen Guest image to bring it back to live. Both the delta consolidation and the freeze itself have a chance to create a consistency or corruption issue in the N-able N-central database. Under most circumstances this chance is low but can be difficult to recover from without an N-able N-central backup to fall back on. Additionally,we do not recommend that VM snapshots are used in short-term schedule (high cost, short retention, high performance for both backup and restore) as these backups can cause performance issues when referencing the delta changes frequently.

This brings us to our best recommendation.

With the tools integrated you can use them to create your schedule. We recommend the process as follows:

1. Shut Down the N-able N-central VM (Shutdown Guest OS, avoid Power Off without the Guest OS shut down);
2. Perform the snapshot;
3. Restart the N-able N-central VM.

This process will provide a polite shutdown of the services and thereby prevent any pause/freeze issues and provide a clean snapshot of the server. We do recommend not keeping them in retention for too long per the Host software recommendations to avoid consolidation issues, but this will be the smoothest snapshot process that will have the best chance of preventing any issues with the data for your server.

Please Note: While a snapshot in this manner before an upgrade or other maintenance to ensure you have a quick and effective way to restore to is worthwhile, if you maintain a snapshot by keeping it enabled the server will suffer from performance issues while delta is in place even when using the above recommendation. You will also need to ensure that you perform a consolidation regularly with the above method to prevent the performance gap from growing. The server will need to refer to both the snapshot and the delta for each transaction to ensure it is using the correct data, which will slow down the server processes. Snapshots daily with the above method can help mitigate (but not prevent) these performance problems as it will consolidate the delta information.