Re: [zfs-macos] block traversal size 819742315008 != alloc 1154012371456 (leaked 334270056448)
Awesome news!!! Jason Belec Sent from my "It's an iPod, a Phone, and an Internet Device..." > On Nov 15, 2014, at 12:12 PM, Simon Casady wrote: > > Thanks for the comments Jason, it apparently was a transient of some kind. > A scrub showed no problems but unmounts and mounts both failed and zdb -b > showed a huge number of leaked space. > > A reboot fixed everything. > > > > > >> On Fri, Nov 14, 2014 at 3:31 PM, BelecMartin >> wrote: >> OK. >> >> Of course 'never' is not a term associated with rotating rust. >> >> I have data that suggests 80% of failures are do to cables, they just take >> drives along for fun. Check the connectors, any split? Sometimes only >> noticeable when warm to hot. >> >> Have you check the power across various points on the supply, motherboard, >> drives? >> >> If your really sure it cannot in any way be hardware, I'm at a loss. I have >> yet to see a single software related failure. >> >> >> >> Jason Belec >> Sent from my "It's an iPod, a Phone, and an Internet Device..." >> >>> On Nov 14, 2014, at 4:22 PM, Simon Casady wrote: >>> >>> These are internal drives and have never been flaky. I use this pool as >>> backup so nothing is lost I just have a need to understand what may have >>> happened before I recreate the pool and how, maybe, to recover. The next >>> time it may be more critical. >>> Looking at console there are crash dumps that show zfs crashing in >>> avl_insert trying to update its mount table, not good. >>> >>> I love poking internals. :) >>> On Fri, Nov 14, 2014 at 3:06 PM, BelecMartin wrote: Do you have snapshots of what you consider 'leaked'? You said a receive failed Have you checked your cables? Then shutdown and bring back online and get status. I've seen lots of weird stuff and 90% can be resolved fairly quickly. You just won't be using every day ZPOOL commands but it is important to have a good understanding of your predicament and you really should be working on a bit for bit clone. I'll help in any way I can and you if you feel like poking the internals. Jason Belec Sent from my "It's an iPod, a Phone, and an Internet Device..." > On Nov 14, 2014, at 3:28 PM, Simon Casady wrote: > > This is from zdb -b on a pool with one vdev of 2 2terabyte drives > mirrored. > > I can trash it and start over but is there away to "unleak" whats leaked ? > I will recreate the pool anyway but would like to experiment with it for > a while first. > > A zfs recv failed with bus error 10 which apparently corrupted something. > zpool status shows no errors. However mount fails for some file systems. > -- > > --- > You received this message because you are subscribed to the Google Groups > "zfs-macos" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to zfs-macos+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. -- --- You received this message because you are subscribed to the Google Groups "zfs-macos" group. To unsubscribe from this group and stop receiving emails from it, send an email to zfs-macos+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout. >>> >>> -- >>> >>> --- >>> You received this message because you are subscribed to the Google Groups >>> "zfs-macos" group. >>> To unsubscribe from this group and stop receiving emails from it, send an >>> email to zfs-macos+unsubscr...@googlegroups.com. >>> For more options, visit https://groups.google.com/d/optout. >> >> -- >> >> --- >> You received this message because you are subscribed to the Google Groups >> "zfs-macos" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to zfs-macos+unsubscr...@googlegroups.com. >> For more options, visit https://groups.google.com/d/optout. > > -- > > --- > You received this message because you are subscribed to the Google Groups > "zfs-macos" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to zfs-macos+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. -- --- You received this message because you are subscribed to the Google Groups "zfs-macos" group. To unsubscribe from this group and stop receiving emails from it, send an email to zfs-macos+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: [zfs-macos] block traversal size 819742315008 != alloc 1154012371456 (leaked 334270056448)
Thanks for the comments Jason, it apparently was a transient of some kind. A scrub showed no problems but unmounts and mounts both failed and zdb -b showed a huge number of leaked space. A reboot fixed everything. On Fri, Nov 14, 2014 at 3:31 PM, BelecMartin wrote: > OK. > > Of course 'never' is not a term associated with rotating rust. > > I have data that suggests 80% of failures are do to cables, they just take > drives along for fun. Check the connectors, any split? Sometimes only > noticeable when warm to hot. > > Have you check the power across various points on the supply, motherboard, > drives? > > If your really sure it cannot in any way be hardware, I'm at a loss. I > have yet to see a single software related failure. > > > > Jason Belec > Sent from my "It's an iPod, a Phone, and an Internet Device..." > > On Nov 14, 2014, at 4:22 PM, Simon Casady wrote: > > These are internal drives and have never been flaky. I use this pool as > backup so nothing is lost I just have a need to understand what may have > happened before I recreate the pool and how, maybe, to recover. The next > time it may be more critical. > Looking at console there are crash dumps that show zfs crashing in > avl_insert trying to update its mount table, not good. > > I love poking internals. :) > > On Fri, Nov 14, 2014 at 3:06 PM, BelecMartin > wrote: > >> Do you have snapshots of what you consider 'leaked'? >> >> You said a receive failed Have you checked your cables? Then shutdown >> and bring back online and get status. >> >> I've seen lots of weird stuff and 90% can be resolved fairly quickly. You >> just won't be using every day ZPOOL commands but it is important to have a >> good understanding of your predicament and you really should be working on >> a bit for bit clone. >> >> I'll help in any way I can and you if you feel like poking the internals. >> >> Jason Belec >> Sent from my "It's an iPod, a Phone, and an Internet Device..." >> >> On Nov 14, 2014, at 3:28 PM, Simon Casady wrote: >> >> This is from zdb -b on a pool with one vdev of 2 2terabyte drives >> mirrored. >> >> I can trash it and start over but is there away to "unleak" whats leaked ? >> I will recreate the pool anyway but would like to experiment with it for >> a while first. >> >> A zfs recv failed with bus error 10 which apparently corrupted something. >> zpool status shows no errors. However mount fails for some file systems. >> >> -- >> >> --- >> You received this message because you are subscribed to the Google Groups >> "zfs-macos" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to zfs-macos+unsubscr...@googlegroups.com. >> For more options, visit https://groups.google.com/d/optout. >> >> -- >> >> --- >> You received this message because you are subscribed to the Google Groups >> "zfs-macos" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to zfs-macos+unsubscr...@googlegroups.com. >> For more options, visit https://groups.google.com/d/optout. >> > > -- > > --- > You received this message because you are subscribed to the Google Groups > "zfs-macos" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to zfs-macos+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. > > -- > > --- > You received this message because you are subscribed to the Google Groups > "zfs-macos" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to zfs-macos+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. > -- --- You received this message because you are subscribed to the Google Groups "zfs-macos" group. To unsubscribe from this group and stop receiving emails from it, send an email to zfs-macos+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: [zfs-macos] block traversal size 819742315008 != alloc 1154012371456 (leaked 334270056448)
OK. Of course 'never' is not a term associated with rotating rust. I have data that suggests 80% of failures are do to cables, they just take drives along for fun. Check the connectors, any split? Sometimes only noticeable when warm to hot. Have you check the power across various points on the supply, motherboard, drives? If your really sure it cannot in any way be hardware, I'm at a loss. I have yet to see a single software related failure. Jason Belec Sent from my "It's an iPod, a Phone, and an Internet Device..." > On Nov 14, 2014, at 4:22 PM, Simon Casady wrote: > > These are internal drives and have never been flaky. I use this pool as > backup so nothing is lost I just have a need to understand what may have > happened before I recreate the pool and how, maybe, to recover. The next > time it may be more critical. > Looking at console there are crash dumps that show zfs crashing in avl_insert > trying to update its mount table, not good. > > I love poking internals. :) > >> On Fri, Nov 14, 2014 at 3:06 PM, BelecMartin >> wrote: >> Do you have snapshots of what you consider 'leaked'? >> >> You said a receive failed Have you checked your cables? Then shutdown >> and bring back online and get status. >> >> I've seen lots of weird stuff and 90% can be resolved fairly quickly. You >> just won't be using every day ZPOOL commands but it is important to have a >> good understanding of your predicament and you really should be working on a >> bit for bit clone. >> >> I'll help in any way I can and you if you feel like poking the internals. >> >> Jason Belec >> Sent from my "It's an iPod, a Phone, and an Internet Device..." >> >>> On Nov 14, 2014, at 3:28 PM, Simon Casady wrote: >>> >>> This is from zdb -b on a pool with one vdev of 2 2terabyte drives mirrored. >>> >>> I can trash it and start over but is there away to "unleak" whats leaked ? >>> I will recreate the pool anyway but would like to experiment with it for a >>> while first. >>> >>> A zfs recv failed with bus error 10 which apparently corrupted something. >>> zpool status shows no errors. However mount fails for some file systems. >>> -- >>> >>> --- >>> You received this message because you are subscribed to the Google Groups >>> "zfs-macos" group. >>> To unsubscribe from this group and stop receiving emails from it, send an >>> email to zfs-macos+unsubscr...@googlegroups.com. >>> For more options, visit https://groups.google.com/d/optout. >> >> -- >> >> --- >> You received this message because you are subscribed to the Google Groups >> "zfs-macos" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to zfs-macos+unsubscr...@googlegroups.com. >> For more options, visit https://groups.google.com/d/optout. > > -- > > --- > You received this message because you are subscribed to the Google Groups > "zfs-macos" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to zfs-macos+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. -- --- You received this message because you are subscribed to the Google Groups "zfs-macos" group. To unsubscribe from this group and stop receiving emails from it, send an email to zfs-macos+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: [zfs-macos] block traversal size 819742315008 != alloc 1154012371456 (leaked 334270056448)
These are internal drives and have never been flaky. I use this pool as backup so nothing is lost I just have a need to understand what may have happened before I recreate the pool and how, maybe, to recover. The next time it may be more critical. Looking at console there are crash dumps that show zfs crashing in avl_insert trying to update its mount table, not good. I love poking internals. :) On Fri, Nov 14, 2014 at 3:06 PM, BelecMartin wrote: > Do you have snapshots of what you consider 'leaked'? > > You said a receive failed Have you checked your cables? Then shutdown > and bring back online and get status. > > I've seen lots of weird stuff and 90% can be resolved fairly quickly. You > just won't be using every day ZPOOL commands but it is important to have a > good understanding of your predicament and you really should be working on > a bit for bit clone. > > I'll help in any way I can and you if you feel like poking the internals. > > Jason Belec > Sent from my "It's an iPod, a Phone, and an Internet Device..." > > On Nov 14, 2014, at 3:28 PM, Simon Casady wrote: > > This is from zdb -b on a pool with one vdev of 2 2terabyte drives mirrored. > > I can trash it and start over but is there away to "unleak" whats leaked ? > I will recreate the pool anyway but would like to experiment with it for > a while first. > > A zfs recv failed with bus error 10 which apparently corrupted something. > zpool status shows no errors. However mount fails for some file systems. > > -- > > --- > You received this message because you are subscribed to the Google Groups > "zfs-macos" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to zfs-macos+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. > > -- > > --- > You received this message because you are subscribed to the Google Groups > "zfs-macos" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to zfs-macos+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. > -- --- You received this message because you are subscribed to the Google Groups "zfs-macos" group. To unsubscribe from this group and stop receiving emails from it, send an email to zfs-macos+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: [zfs-macos] block traversal size 819742315008 != alloc 1154012371456 (leaked 334270056448)
Do you have snapshots of what you consider 'leaked'? You said a receive failed Have you checked your cables? Then shutdown and bring back online and get status. I've seen lots of weird stuff and 90% can be resolved fairly quickly. You just won't be using every day ZPOOL commands but it is important to have a good understanding of your predicament and you really should be working on a bit for bit clone. I'll help in any way I can and you if you feel like poking the internals. Jason Belec Sent from my "It's an iPod, a Phone, and an Internet Device..." > On Nov 14, 2014, at 3:28 PM, Simon Casady wrote: > > This is from zdb -b on a pool with one vdev of 2 2terabyte drives mirrored. > > I can trash it and start over but is there away to "unleak" whats leaked ? > I will recreate the pool anyway but would like to experiment with it for a > while first. > > A zfs recv failed with bus error 10 which apparently corrupted something. > zpool status shows no errors. However mount fails for some file systems. > -- > > --- > You received this message because you are subscribed to the Google Groups > "zfs-macos" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to zfs-macos+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. -- --- You received this message because you are subscribed to the Google Groups "zfs-macos" group. To unsubscribe from this group and stop receiving emails from it, send an email to zfs-macos+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
