I have a situation where my zpool (with two radiz2s) is resilvering and reaches a certain point, then starts over. There no read, write or checksum errors. The disks do have a fair amount of resilvering to do, as I''ve had a variety of disk failures. But at the core of things, there''s enough parity data to avoid data loss, and the zpool is functioning fine, albeit in a degraded state. I''ve been trapped in this loop before and it''s been resolved by changing the disk configuration somehow. But that is just like shooting in the dark and can be a very time-consuming process. Is there a more proper way to approach this issue? Should I be filing a bug report? -Galen
Galen wrote:> I have a situation where my zpool (with two radiz2s) is resilvering > and reaches a certain point, then starts over. > > There no read, write or checksum errors. The disks do have a fair > amount of resilvering to do, as I''ve had a variety of disk failures. > But at the core of things, there''s enough parity data to avoid data > loss, and the zpool is functioning fine, albeit in a degraded state. > > I''ve been trapped in this loop before and it''s been resolved by > changing the disk configuration somehow. But that is just like > shooting in the dark and can be a very time-consuming process. > > Is there a more proper way to approach this issue? Should I be filing > a bug report?What release and/or build? -- Andrew
I''ve never used snapshotting on this zpool and I can''t imagine how it would snapshot without my involvement. -Galen On Jul 11, 2009, at 2:22 PM, Thomas Burgess wrote:> do you have auto-snapshots on? i think old versions would restart on > every single new snapshot > > > On Sat, Jul 11, 2009 at 5:19 PM, Galen <galenz at zinkconsulting.com> > wrote: > I have a situation where my zpool (with two radiz2s) is resilvering > and reaches a certain point, then starts over. > > There no read, write or checksum errors. The disks do have a fair > amount of resilvering to do, as I''ve had a variety of disk failures. > But at the core of things, there''s enough parity data to avoid data > loss, and the zpool is functioning fine, albeit in a degraded state. > > I''ve been trapped in this loop before and it''s been resolved by > changing the disk configuration somehow. But that is just like > shooting in the dark and can be a very time-consuming process. > > Is there a more proper way to approach this issue? Should I be > filing a bug report? > > -Galen > _______________________________________________ > zfs-discuss mailing list > zfs-discuss at opensolaris.org > http://mail.opensolaris.org/mailman/listinfo/zfs-discuss >
On Jul 11, 2009, at 2:24 PM, Andrew Gabriel wrote:> Galen wrote: >> I have a situation where my zpool (with two radiz2s) is resilvering >> and reaches a certain point, then starts over. >> >> There no read, write or checksum errors. The disks do have a fair >> amount of resilvering to do, as I''ve had a variety of disk >> failures. But at the core of things, there''s enough parity data to >> avoid data loss, and the zpool is functioning fine, albeit in a >> degraded state. >> >> I''ve been trapped in this loop before and it''s been resolved by >> changing the disk configuration somehow. But that is just like >> shooting in the dark and can be a very time-consuming process. >> >> Is there a more proper way to approach this issue? Should I be >> filing a bug report? > > What release and/or build? > > -- > AndrewOpenSolaris 2009.06, all updates applied, pretty much clean install.
On Jul 11, 2009, at 5:41 PM, Galen <galenz at zinkconsulting.com> wrote:> > On Jul 11, 2009, at 2:24 PM, Andrew Gabriel wrote: > >> Galen wrote: >>> I have a situation where my zpool (with two radiz2s) is >>> resilvering and reaches a certain point, then starts over. >>> >>> There no read, write or checksum errors. The disks do have a fair >>> amount of resilvering to do, as I''ve had a variety of disk >>> failures. But at the core of things, there''s enough parity data to >>> avoid data loss, and the zpool is functioning fine, albeit in a >>> degraded state. >>> >>> I''ve been trapped in this loop before and it''s been resolved by >>> changing the disk configuration somehow. But that is just like >>> shooting in the dark and can be a very time-consuming process. >>> >>> Is there a more proper way to approach this issue? Should I be >>> filing a bug report? >> >> What release and/or build? >> >> -- >> Andrew > > OpenSolaris 2009.06, all updates applied, pretty much clean install.Did you make sure the auto snapshot service was disabled? Otherwise it will be taking hourly snapshots. I think it''s enabled by default. -Ross
On Jul 11, 2009, at 4:00 PM, Ross Walker wrote:> On Jul 11, 2009, at 5:41 PM, Galen <galenz at zinkconsulting.com> wrote: > >> >> On Jul 11, 2009, at 2:24 PM, Andrew Gabriel wrote: >> >>> Galen wrote: >>>> I have a situation where my zpool (with two radiz2s) is >>>> resilvering and reaches a certain point, then starts over. >>>> >>>> There no read, write or checksum errors. The disks do have a fair >>>> amount of resilvering to do, as I''ve had a variety of disk >>>> failures. But at the core of things, there''s enough parity data >>>> to avoid data loss, and the zpool is functioning fine, albeit in >>>> a degraded state. >>>> >>>> I''ve been trapped in this loop before and it''s been resolved by >>>> changing the disk configuration somehow. But that is just like >>>> shooting in the dark and can be a very time-consuming process. >>>> >>>> Is there a more proper way to approach this issue? Should I be >>>> filing a bug report? >>> >>> What release and/or build? >>> >>> -- >>> Andrew >> >> OpenSolaris 2009.06, all updates applied, pretty much clean install. > > Did you make sure the auto snapshot service was disabled? Otherwise > it will be taking hourly snapshots. > > I think it''s enabled by default. > > -Ross >There''s no snapshotting happening to my knowledge. Moreover, if I change the disk configuration around enough, it will resilver without issues. -Galen