Archives
Archives
| From: | Julien Roy <julien@××××.ca> | ||
|---|---|---|---|
| To: | Gentoo User <gentoo-user@l.g.o> | ||
| Subject: | Re: [gentoo-user] Hard drive error from SMART | ||
| Date: | Tue, 12 Apr 2022 22:17:02 | ||
| Message-Id: | N-UqRPH--3-2@jroy.ca | ||
| In Reply to: | Re: [gentoo-user] Hard drive error from SMART by Wol |
||
| 1 | > For those people looking at btrfs - note that parity-raid (5 or 6) is not a wise idea at the moment so you don't get two-failure protection ... |
| 2 | > |
| 3 | > Cheers, |
| 4 | > Wol |
| 5 | > |
| 6 | I've been reading that this is less and less true. The write-hole issue is rather old now (first reported around 2016 I think?) From what I read from various sources, the developpers have made some progress and the problem is getting harder and harder to reproduce, for instance, [1]. |
| 7 | Although some people recommend using RAID1 for the metadata, and RAID5/6 for the data, just in case. |
| 8 | |
| 9 | |
| 10 | Julien |
| 11 | [1] https://unixsheikh.com/articles/battle-testing-zfs-btrfs-and-mdadm-dm.html#btrfs-raid-5 |