# dumpfs /dev/rwd1l | head -1 magic 11954 (FFS1) time Wed Jan 8 19:45:37 2020 # dumpfs /dev/rwd2l | head -1 magic 11954 (FFS1) time Sun Mar 27 13:01:58 2022
On Tue, Mar 29, 2022 at 11:07 AM Otto Moerbeek <o...@drijf.net> wrote: > On Tue, Mar 29, 2022 at 10:58:49AM -0400, F Bax wrote: > > > I used rsync to copy files. > > sudo rsync -anv --delete /mnt/wd1l/ /mnt/wd2l/ > > reports no changes required (runtime under 3 minutes). > > sudo diff -r /mnt/wd1l/ /mnt/wd2l/ > > reports no difference (runtime 10 hours) > > > > $ sudo df -i /mnt/wd1l/ /mnt/wd2l/ > > Filesystem 512-blocks Used Avail Capacity iused ifree %iused > > Mounted on > > /dev/wd1l 2138940784 1997329632 34664128 98% 483707 33313411 > 1% > > /mnt/wd1l > > /dev/wd2l 2138951776 2033043696 -1039504 100% 483707 33313411 > 1% > > /mnt/wd2l > > Ok, then it could be an FFS1 vs FFS2 thing. FFS2 has a larger > meta-data overhead due to its larger inodes. > > Check > > # dumpfs /dev/rwd1l | head -1 > # dumpfs /dev/rwd2l | head -1 > > -Otto > > > > > On Tue, Mar 29, 2022 at 10:49 AM F Bax <fbax...@gmail.com> wrote: > > > > > I used rsync to copy files. df -i reports 483707 inodes used for both > > > partitions. > > > sudo rsync -anv --delete /mnt/wd1l/ /mnt/wd2l/ > > > reports no changes required (runtime under 3 minutes). > > > sudo diff -r /mnt/wd1l/ /mnt/wd2l/ > > > reports no difference (runtime 10 hours) > > > > > > On Tue, Mar 29, 2022 at 10:39 AM Otto Moerbeek <o...@drijf.net> wrote: > > > > > >> On Tue, Mar 29, 2022 at 10:25:34AM -0400, F Bax wrote: > > >> > > >> > I copied all files from /mnt/wd1l to /mnt/wd2l > > >> > > > >> > wd2l is slightly larger than wd1l; yet wd2l is full! > > >> > > > >> > $ df -h /mnt/wd1l /mnt/wd2l > > >> > Filesystem Size Used Avail Capacity Mounted on > > >> > /dev/wd1l 1020G 952G 16.5G 98% /mnt/wd1l > > >> > /dev/wd2l 1020G 969G -508M 100% /mnt/wd2l > > >> > > >> How did you copy? Some forms of copy will cause hardlinked files to be > > >> separate files on the destination. df -i will tell how many inodes you > > >> have used. If wd2l has more inodes in use, I bet it's that. > > >> > > >> -Otto > > >> > > >> > > > >> > Output from disklabel is almost identical: > > >> > > > >> > type: SCSI > > >> > disk: SCSI disk > > >> > label: WDC WD2000FYYZ-0 > > >> > flags: > > >> > bytes/sector: 512 > > >> > sectors/track: 63 > > >> > tracks/cylinder: 255 > > >> > sectors/cylinder: 16065 > > >> > cylinders: 243201 > > >> > total sectors: 3907029168 > > >> > rpm: 0 > > >> > interleave: 1 > > >> > trackskew: 0 > > >> > cylinderskew: 0 > > >> > headswitch: 0 # microseconds > > >> > track-to-track seek: 3907029168 # microseconds > > >> > drivedata: 0 > > >> > > > >> > Difference between wd1 and wd2: > > >> > wd1: interleave: 0 > > >> > wd2: interleave: 1 > > >> > > > >> > Partition details (A added 'wd1/wd2' to beginning of line: > > >> > # size offset fstype [fsize bsize cpg] > > >> > wd1l: 2147472640 525486208 4.2BSD 8192 65536 1 > > >> > wd2l: 2147483647 63 4.2BSD 8192 65536 1 > > >> > > > >> > Why is wd2l full? > > >> > > > >
