On Wed, Mar 03, 2021 at 05:16:14PM -0500, Felix Miata wrote: > Andrei POPESCU composed on 2021-03-03 17:50 (UTC+0200): > > > Felix Miata wrote: > > >> To start with, RAID1 is marginally slower than ordinary filesystems on partitions. > > > This is true for some workloads, for others it can be significantly > > faster. > > > https://arstechnica.com/information-technology/2020/04/understanding-raid-how-performance-scales-from-one-disk-to-eight/ > > [...] Do > you know of, or can you provide a reference to, any way RAID1 performance can be > better than single disk? I can just provide a hint: when accessing mirrored disks, reading is done whenever the first disk finishes; writing whenever both disks are done. So reading is potentially faster, writing potentially slower. That's about latency. The throughput shouldn't change too much (unless you're combining very different disks, but that'd be asking for it). Compose this with several layers of buffering and I doubt you'll be able to make out any significant difference though. No experimental data to back this up, alas. > Evolution as taught in public schools, like religion, > is based on faith, not on science. OK. Now all bets wrt RAID are off, anyway. Cheers - t
Attachment:
signature.asc
Description: Digital signature