--- Ursprüngliche Nachricht --- Von: Goffredo Baroncelli <kreijack@xxxxxxxxx> Datum: 04.02.2021 19:13:50 An: Cedric.dewijs@xxxxxxxxxx, Andy Smith <andy@xxxxxxxxxxxxxx> Betreff: Re: put 2 hard drives in mdadm raid 1 and detect bitrot like btrfs does, what's that called? [...] > Hey Andy, > > I would rather see performance figures for these setups: > A) btrfs with 2 (or more) hard drives and one SSD in writeback bcache configuration (unsafe against failure of the ssd): > +-----------------------------+ > | btrfs raid 1 /mnt | > +--------------+--------------+ > | /dev/Bcache0 | /dev/Bcache1 | > +--------------+--------------+ > | bcache writeback Cache | > | /dev/sdk1 | > +--------------+--------------+ > | Data | Data | > | /dev/sdv1 | /dev/sdw1 | > +--------------+--------------+ Doing that, you loose the protection of raid1 redundancy: now there is a single point of failure /dev/sdk1. Writeback is even more dangerous... Not really. if bcache is set to read cache, the SSD can die at any moment, without btrfs loosing any data. All written data has gone straight to the hard drives. I have not tried this scenario, but I would be very surprised if reading the data from /mnt is even interrupted for longer than a few seconds if the data cable from the ssd is pulled while data is written from another process. You are correct about writeback cache, if /dev/sdk1 dies, all dirty data is lost, and even worse, both copies of the btrfs data are side by side on only the SSD. (But I already mentioned this in my previous mail: "unsafe against failure of the ssd") Cheers, Cedric --- Take your mailboxes with you. Free, fast and secure Mail & Cloud: https://www.eclipso.eu - Time to change!
