David Dyer-Bennet
2009-Feb-16 05:02 UTC
[zfs-discuss] Can''t find results of send/receive on destination fs??
So I did this: zfs send -R "rpool/export/home at bup-20090216-044512UTC" | zfs receive\ -dv "bup-ruin/fsfs" And it indicated no errors, and said it was creating various expected filesystems (pool bup-ruin was brand new at this test). And afterward bup-ruin seemed to have reasonable filesystems in it: localddb at fsfs:~/src/bup2# zfs list -r bup-ruin NAME USED AVAIL REFER MOUNTPOINT bup-ruin 35.2M 913G 23K /backups/bup-ruin bup-ruin/fsfs 35.0M 913G 18K /backups/bup-ruin/fsfs bup-ruin/fsfs/export 35.0M 913G 18K /backups/bup-ruin/fsfs/export bup-ruin/fsfs/export/home 35.0M 913G 19K /backups/bup-ruin/export/home bup-ruin/fsfs/export/home/localddb 35.0M 913G 27.8M /backups/bup-ruin/export/home/loca\ lddb localddb at fsfs:~/src/bup2# zfs list -t snapshot -r bup-ruin/fsfs/export NAME USED AVAIL REFER MOUNTPOINT bup-ruin/fsfs/export/home at bup-20090216-044512UTC 0 - 19K - bup-ruin/fsfs/export/home/localddb at first 26K - 27.6M - bup-ruin/fsfs/export/home/localddb at second 28K - 27.6M - bup-ruin/fsfs/export/home/localddb at bup-20090216-044512UTC 0 - 27.8M - And, checking a bit more deeply, the snapshot created for this backup is there, and two previous snapshots are there. Or claim to be there. But I can''t actually find them anywhere. localddb at fsfs:~/src/bup2# pfexec /usr/bin/ls -a /backups/bup-ruin/fsfs/export . .. localddb at fsfs:~/src/bup2# pfexec /usr/bin/ls -a /backups/bup-ruin/fsfs/export/.zfs . .. snapshot localddb at fsfs:~/src/bup2# pfexec /usr/bin/ls -a /backups/bup-ruin/fsfs/export/.zfs/snapshot\ / . .. Note that bup-ruin is mounted with an altroot property: localddb at fsfs:~/src/bup2# zpool list NAME SIZE USED AVAIL CAP HEALTH ALTROOT bup-ruin 928G 35.3M 928G 0% ONLINE /backups/bup-ruin rpool 74G 3.85G 70.1G 5% ONLINE - zp1 744G 451G 293G 60% ONLINE - I think I need this, because the filesystems being backed up onto it have mountpoints of their own, and the originals are currently occupying those. I''m about 100.0% sure this is user error (me being the user), but I''m stuck on this right now. What am I expecting wrong? Or doing wrong? -- David Dyer-Bennet, dd-b at dd-b.net; http://dd-b.net/ Snapshots: http://dd-b.net/dd-b/SnapshotAlbum/data/ Photos: http://dd-b.net/photography/gallery/ Dragaera: http://dragaera.info
Frank Cusack
2009-Feb-17 03:31 UTC
[zfs-discuss] Can''t find results of send/receive on destination fs??
On February 15, 2009 11:02:51 PM -0600 David Dyer-Bennet <dd-b at dd-b.net> wrote:> So I did this: > > zfs send -R "rpool/export/home at bup-20090216-044512UTC" | zfs receive\ > -dv "bup-ruin/fsfs" > > And it indicated no errors, and said it was creating various expected > filesystems (pool bup-ruin was brand new at this test)....> Note that bup-ruin is mounted with an altroot property:...> I think I need this, because the filesystems being backed up onto it have > mountpoints of their own, and the originals are currently occupying those.You do.> I''m about 100.0% sure this is user error (me being the user), but I''m > stuck on this right now. What am I expecting wrong? Or doing wrong?Works for me, Solaris 10 U6 x86. -frank
David Dyer-Bennet
2009-Feb-17 03:42 UTC
[zfs-discuss] Can''t find results of send/receive on destination fs??
On Mon, February 16, 2009 21:31, Frank Cusack wrote:> On February 15, 2009 11:02:51 PM -0600 David Dyer-Bennet <dd-b at dd-b.net> > wrote: >> So I did this: >> >> zfs send -R "rpool/export/home at bup-20090216-044512UTC" | zfs receive\ >> -dv "bup-ruin/fsfs" >> >> And it indicated no errors, and said it was creating various expected >> filesystems (pool bup-ruin was brand new at this test). > ... >> Note that bup-ruin is mounted with an altroot property: > ... >> I think I need this, because the filesystems being backed up onto it >> have >> mountpoints of their own, and the originals are currently occupying >> those. > > You do.Finally made a simple enough situation that I could find it by brute force. They were up and over a level from where I expected them to be, *and* there were empty filesystems where I expected them to be, which confused things mightily. But brute force searching for a file eventually turned it up. In hindsight, they''re exactly where they should have been :-). -- David Dyer-Bennet, dd-b at dd-b.net; http://dd-b.net/ Snapshots: http://dd-b.net/dd-b/SnapshotAlbum/data/ Photos: http://dd-b.net/photography/gallery/ Dragaera: http://dragaera.info