BLU Discuss list archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Discuss] ZFS error checking question
- Subject: [Discuss] ZFS error checking question
- From: richard.pieri at gmail.com (Rich Pieri)
- Date: Fri, 5 Aug 2022 20:21:29 -0400
- In-reply-to: <CAJFsZ=qKUzJ+WydgJszxArrBA7qcdLgtNogcLjd8_hWy-M=2TQ@mail.gmail.com>
- References: <0f9cacf3-e355-1109-6e92-2cab27643886@borg.org> <20220725115851.5f73ed89.Richard.Pieri@gmail.com> <CAJFsZ=qKUzJ+WydgJszxArrBA7qcdLgtNogcLjd8_hWy-M=2TQ@mail.gmail.com>
On Fri, 5 Aug 2022 18:31:40 -0400 Bill Bogstad <bogstad at pobox.com> wrote: > Assuming for the moment that ZFS is talking directly to a storage > device rather than some kind of virtualized environment, how does ZFS > manage to avoid just getting back what might be in the RAM buffers of > the device rather > than reaching all the way down to the physical media (be it either > spinning rust or > flash cells)? It can't. When ZFS sends the flush signal and the drive returns confirmation, ZFS expects the truth. If the drive lies then all bets are off. All consumer class drives lie and some enterprise class drives lie. -- \m/ (--) \m/
- References:
- [Discuss] ZFS error checking question
- From: bogstad at pobox.com (Bill Bogstad)
- [Discuss] ZFS error checking question
- Prev by Date: [Discuss] ZFS error checking question
- Next by Date: [Discuss] Boston Linux VIRTUAL Meeting , Wednesday, August 17, 2022 - ZFS and the Infinite Incremental Backup
- Previous by thread: [Discuss] ZFS error checking question
- Next by thread: [Discuss] Boston Linux VIRTUAL Meeting , Wednesday, August 17, 2022 - ZFS and the Infinite Incremental Backup
- Index(es):