Hello, I'm trying to replicate an entire pool onto another pool in a different server, a full back-up of the entire pool including everything (pull). In the past, I've just checked the "(Almost) full filesystem replication" as well as checking all the datasets in the source selector, and it's worked. However at some point it's started failing with errors relating to the two zvols that exist among the data, likely when I updated to 12.0U5, though I'm not entirely sure if there's a correlation. Both systems are currently running 12.0U6.
I'm attaching some logs and screenshots to this post. These lines are present with every failure:
Sometimes it is followed by this:
As a workaround right now, I'm running the backup without having the top-level dataset selected, which seems to allow it to run without errors, replicating all datasets and zvols just fine, though of course without the snapshots of the top-level dataset. There's no data other than child datasets directly on tank, so I don't think I'm losing any data by doing it this way, but it seems like an ugly workaround given that my purpose is to have two servers with absolutely identical data on them.
Any thoughts on this?
I'm attaching some logs and screenshots to this post. These lines are present with every failure:
Code:
Warning: tank/iSCSI-devs/workstation-disk: property 'sharesmb' does not apply to datasets of this type Warning: tank/iSCSI-devs/workstation-disk: property 'mountpoint' does not apply to datasets of this type Warning: tank/iSCSI-devs/workstation-disk: property 'sharenfs' does not apply to datasets of this type Warning: tank/iSCSI-devs/iis-storage: property 'sharesmb' does not apply to datasets of this type Warning: tank/iSCSI-devs/iis-storage: property 'mountpoint' does not apply to datasets of this type Warning: tank/iSCSI-devs/iis-storage: property 'sharenfs' does not apply to datasets of this type destination 'tank/.system/rrd-1e6cbdfb415748b98d868947a6e14a88' does not exist Command failed with code 1.
Sometimes it is followed by this:
Code:
Broken pipe. The snapshot tank/.system@2021-10-28_06-00-TANK-AUTO-1W was not transferred. Please run `zfs destroy tank@2021-10-28_06-00-TANK-AUTO-1W` on the target system and run replication again.
As a workaround right now, I'm running the backup without having the top-level dataset selected, which seems to allow it to run without errors, replicating all datasets and zvols just fine, though of course without the snapshots of the top-level dataset. There's no data other than child datasets directly on tank, so I don't think I'm losing any data by doing it this way, but it seems like an ugly workaround given that my purpose is to have two servers with absolutely identical data on them.
Any thoughts on this?
