XCP-ng
    • Categories
    • Recent
    • Tags
    • Popular
    • Users
    • Groups
    • Register
    • Login

    Replication retention & max chain size

    Scheduled Pinned Locked Moved Backup
    9 Posts 3 Posters 369 Views 3 Watching
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • M Offline
      McHenry
      last edited by McHenry

      If using continuous replication am I correct in understanding the maximum chain length is 30 elements so the maximum retention vaue should be the same?

      https://xen-orchestra.com/docs/backup_troubleshooting.html#vdi-chain-protection

      1 Reply Last reply Reply Quote 0
      • olivierlambertO Online
        olivierlambert Vates 🪐 Co-Founder CEO
        last edited by

        Yes, that's correct. Replication isn't meant to have high retention BTW, if you want that, it's more like backup that you need.

        M 1 Reply Last reply Reply Quote 0
        • M Offline
          McHenry @olivierlambert
          last edited by McHenry

          @olivierlambert

          High retention meaning replication should not be used for meaning long storage term?

          As a proof of concept I am looking to use CR for hourly, daily & weekly. Max retention being 4 weeks. Then monthly offsite to Wasabi, possibly using mirror but have not got my head that far yet.

          A 1 Reply Last reply Reply Quote 0
          • olivierlambertO Online
            olivierlambert Vates 🪐 Co-Founder CEO
            last edited by

            Long term is backup, not replication. You can use your Wasabi as long term as a chained backup after the replication.

            1 Reply Last reply Reply Quote 0
            • A Offline
              Andrew Top contributor @McHenry
              last edited by Andrew

              @McHenry As an example.... I use CR hourly for the main pool onto a backup host that has lots of storage and the same networks. It's a quick DR server that can start an VM (or a clone) immediately if the main pool fails or a VM becomes so corrupt that it needs to be restored from backup. But this is not a long term backup solution, it's an immediate restore solution. CR does an hourly delta update and a weekly full update just to make sure the data is always clean. You can think of CR as an off-host snapshot, and you don't want to keep too many around. To be clear, I don't keep ongoing deltas long term, just the current one.

              For long term backup I use Wasabi S3 and do delta backups every night. I keep the deltas for about 90 days and do a new full backup about every 3 months. This again is to make sure the delta backup chain does not have a long term corruption. It is possible to restore any VM or any file from any delta (or full) backup along the way.

              M 2 Replies Last reply Reply Quote 1
              • M Offline
                McHenry @Andrew
                last edited by McHenry

                @Andrew

                Excellent answer, thank you. I am looking at a very similar setup. If I have understood correctly, I understand your setup to be:

                DR Host

                1. Hourly CR - Retention=84 (assuming 12 per day)
                2. Weekly full - Retention=1?

                I have guessed your DR host weekly retention of 1 as I expect this would keep a rolling 7 day cycle?

                Wasabi Remote

                1. Daily Delta - Retention=90
                2. Quarterly full - Retention=?

                For the Wasabi remote could I do:

                1. Daily delta - retention = 7
                2. Weekly delta - retention = 4
                3. Monthly delta - retention = 3
                4. Quarterly full - retention = 4

                Would this result in a full offsite only once every 3 months and everything else would be a delta?

                With the Wasabi remote can this also have a health check to ensure the backup works?

                p.s. I had understood from the above that the max chain length was 30?

                1 Reply Last reply Reply Quote 0
                • M Offline
                  McHenry @Andrew
                  last edited by

                  @Andrew said in Replication retention & max chain size:

                  It's a quick DR server that can start an VM (or a clone) immediately if the main pool fails

                  Does this mean you do not have your production server(s) and DR server in the same pool?

                  At each client site we have a single production server and a single DR server. Both have been added to the same client specific pool however I have no idea if this is good or bad.

                  A 1 Reply Last reply Reply Quote 0
                  • A Offline
                    Andrew Top contributor @McHenry
                    last edited by

                    @McHenry Correct, my DR server is not part of the main pool. The main pool has several hosts and shared storage (NFS) and if one host fails there's room for guest VMs on another host in the pool. If main storage totally fails then the VMs could be run on the DR system and then migrated back to the main pool when operational.

                    The DR server is a form of an active backup as it does not share storage with the main pool (it has its own SSD RAID6). There are other real backups (offsite S3 and separate OS level backups).

                    M 1 Reply Last reply Reply Quote 1
                    • M Offline
                      McHenry @Andrew
                      last edited by

                      @Andrew

                      Got it. As we only have a single production server there is no shared storage so I guess the idea of a pool is mute.

                      1 Reply Last reply Reply Quote 0
                      • First post
                        Last post