On Fri, Sep 11, 2015 at 03:09:21PM +0200, Lars Ellenberg wrote: > On Thu, Sep 10, 2015 at 02:21:02PM +0200, Joerg May wrote: > > > Would e.g. checksum-based syncing help to identify faster the free > > blocks and skip them for syncing ? > > No.
IFF: - the rest (see later) of the source partition contains all zeros (or some other defined value) AND - the partition on you new disk contains the same value AND - your network is the bottleneck and not your CPU (e.g., overhead for calculating checksums). then it might help. Lots of ifs... For example you can not assume that your new disk contains all zeros, there might be test-patterns from the factory and what not. Or you usually can not assume that your old partition contains all data at the beginning and the rest is "empty", there might be holes not used by the file system that contain old garbage. How do you then fill only the holes with the defined value? So, as Lars said, in the usual case no, under certain conditions, yes. -> truck based sync. Regards, rck _______________________________________________ drbd-user mailing list [email protected] http://lists.linbit.com/mailman/listinfo/drbd-user
