Hello,
I have been having this issue with ZFS replication from one machine to another over a 100 mbps link.
Source machine acts as our back up server siphoning data via rsync from other Linux machines and then takes a snapshot of the data each night.
Source machine is a 6 core AMD with 16GB of RAM, 3X3TB WD Reds in RAIDZ1. Destination machine is a dual core AMD with 8GB of ram and 3x2TB drives (source machine holds 1.2TB of data at any given time), intel lan on both machines.
ZFS snapshots are kept for 6 months and then discarded. Source machine runs scrub jobs every Sunday.
The error message i'm getting is: "cannot receive incremental stream: most recent snapshot of remote/dataset does not match incremental source".
When the failure occurs, i check the button for "Initialize remote side for once. (May cause data loss on remote side!):" which causes replication to delete remote data and start fresh, but the same error occurs after a while.
Issues i've noticed during replication would be: source machine is unresponsive or takes minutes for web gui or ssh to allow me to log in, sometimes the web gui fails to allow me to access it with some weird error message.
i'm also running 2 samba shares off the source machine, each of 30 and 40 GB of data.
Those shares are almost impossible to access while replication is in progress.
I have been having this issue with ZFS replication from one machine to another over a 100 mbps link.
Source machine acts as our back up server siphoning data via rsync from other Linux machines and then takes a snapshot of the data each night.
Source machine is a 6 core AMD with 16GB of RAM, 3X3TB WD Reds in RAIDZ1. Destination machine is a dual core AMD with 8GB of ram and 3x2TB drives (source machine holds 1.2TB of data at any given time), intel lan on both machines.
ZFS snapshots are kept for 6 months and then discarded. Source machine runs scrub jobs every Sunday.
The error message i'm getting is: "cannot receive incremental stream: most recent snapshot of remote/dataset does not match incremental source".
When the failure occurs, i check the button for "Initialize remote side for once. (May cause data loss on remote side!):" which causes replication to delete remote data and start fresh, but the same error occurs after a while.
Issues i've noticed during replication would be: source machine is unresponsive or takes minutes for web gui or ssh to allow me to log in, sometimes the web gui fails to allow me to access it with some weird error message.
i'm also running 2 samba shares off the source machine, each of 30 and 40 GB of data.
Those shares are almost impossible to access while replication is in progress.