VERITAS Volume Manager 3.1 Administrator's Guide

Volume Manager Operations
FastResync (Fast Mirror Resynchronization)
Chapter 3118
is used to reattach the snapshot plex. If FMR is enabled before the
snapshot is taken and is not disabled at any time before the snapshot is
complete, then the FMR delta changes reflected in the FMR bitmap are
used to resynchronize the volume during the snapback. To make it easier
to create snapshots of several volumes at the same time, the snapshot
option has been enhanced to accept more than one volume and a naming
scheme has been added. By default, each replica volume is named
SNAP-< original volume >. This default can be overridden with options
on the command line. Only volumes in the same disk group can take a
snapshot of more than one volume at a time.
To make it easier to snapshot all the volumes in a single disk group, the
option -o resyncfromreplica has been added to vxassist. However, it
fails if any of the volumes in the disk group do not have a complete
snapshot plex. It is possible to take several snapshots of the same
volume. A new FMR bitmap is produced for each snapshot taken, and the
resynchronization time for each snapshot is minimized.
The snapshot plex can be chosen as the preferred set of data when
performing a snapback. Adding -o resyncfromreplica to the snapback
option copies the data on the snapshot (replica) plex onto all the mirrors
attached to the original volume. By default, the data on the original
volume is preferred and copied onto the snapshot plex.
It is possible to grow the replica volume, or the original volume, and still
use FMR. Growing the volume extends the bitmap FMR uses to track the
delta changes. This may change the size of the bitmap or its region size.
In either case, the part of the bitmap that corresponds to the grown area
of the volume is marked as “dirty” so that this area is resynchronized.
The snapshot operation fails if the snapshot attempts to create an
incomplete snapshot plex. In such cases, it is necessary to grow the
replica volume, or the original volume, before the snapback option is run.
Growing the two volumes separately can lead to a snapshot that shares
physical disks with another mirror in the volume. To prevent this, grow
the volume after the snapback command is complete.
Any operation that changes the layout of the replica volume can mark
the FMR map for that snapshot “dirty” and require a full
resynchronization during the snapback. Operations that cause this
include subdisk split, subdisk move, and online relayout of the replica. It
is safe to perform these operations after the snapshot is completed. For
more information, see the vxvol (1M), vxassist (1M), and vxplex (1M)
manual pages.