• BackupFan
    We are having an issue with our cloud backups on a customer's server. It has been a few weeks now since the first full backup completed for this server, and a pattern seems to be developing. We have two backup plans: one local and one cloud. Both plans are set to backup daily, and to do a full backup once a week. The pattern seems to be that the cloud backup runs successfully for about 4 days out of the week, but the weekly full backup causes it to take about 3 days to complete. This full backup seems to be causing the daily local backup to fail, since only one plan can run at a time.
    What can we do to ensure that the cloud weekly full backups do not cause local backups to fail?
    It is our understanding that hybrid backups cannot be used with synthetic full backups. Is this correct?
    Is there a way to chain several different backups together, where backup plan A's completion triggers the start of backup plan B?
    Any other tips or information on best practices for backup schedules would be much appreciated.
    Thank you.
  • David Gugick
    Backup Chaining is supported and available in the backup wizard. You are correct that synthetic fulls are not supported with local storage (and, consequently, with hybrid backups).

    I think we need to figure out why the fulls are taking so long to complete. You could reduce your full backup interval to something longer than a week (14 - 28 days) to reduce their frequency. But a support case would likely yield the fastest analysis and possible resolution.

    Can you describe how many systems are being backed up using synthetic full, whether those machines are all running full backups at the same time, which storage you are using (S3?), and how much upstream bandwidth you have?
  • BackupFan
    On that particular network, about three machines are running synthetic backups, all at the same time. However, the machine in question has by far the largest amount of data. We are using Wasabi for our cloud storage, and the upstream bandwidth on this machine is 11 Mbps.
  • David Gugick
    Upstream is slow, unfortunately. I would try running the big full backup without other fulls running at the same time. It's not clear how much data is being uploaded with the synthetic, but it could be relatively low compared to the full size. In that case the issue may be on the API processing for the synthetic rather than bandwidth. I'd encourage you to open a support case so the team can review the logs and figure out where they delay is in processing.
  • BackupFan
    Thank you, I'll look into that.

    I have another question about backup chaining. If backup plan A is scheduled to run a full backup once every two weeks and every other day runs an incremental backup, and backup plan B has no schedule and is only triggered by backup A, how can backup B be told to run periodic full backups? Will backup B run a full backup because backup A was scheduled to do so? I just want to make sure that using backup chaining doesn't mean that I'm setting up one of my backup plans to never run full backups.
    Thank you for all of your help so far!
  • David Gugick
    Are these image backups? I'm assuming so since we do support mutliple backups running concurrently for file backups or a single image backup and multiple file backups.

    I think we need to address the 3 Days for a full to complete before resorting to work-arounds. I need you to open a support case from the agent that is taking 3 days to complete. Use the Tools - Diagnostic option so logs are submitted. The team can review and provide some guidance on the underlying full backup speed issue.

    But, a work-around might be to schedule Backup B with only a Full Backup Schedule (Advanced Scheduling) and let the chain from Backup A kick off the block-level incrementals. I haven't tested this myself, but I believe it would work. I am checking with Support to get clarification on this.

    But please open a support case to discuss the slow synthetic full issue.
  • BackupFan
    That's right, they are image backups.
    I have just opened up a support case and sent the logs from the device in question.

    Thank you for checking on that, I would be very interested to know if that is a viable solution.
  • David Gugick
    Yes, that would be a solution. The chained backup will always run an incremental (unless the force a full option is selected on the chained backup). So running the full as a part of the main backup plan schedule and leaving the chained execution to run the incrementals should work.
Add a Comment

Welcome to MSP360 (CloudBerry) Forum!

Thank you for visiting! Please take a moment to register so that you can participate in the discussions!