Posted by: Anonymous
[ip: 192.168.0.139]
on April 12, 2008 08:56 PM
RAID10 rules for most applications, especially Linux 'md', albeit RAID5 may offer more usable diskspace and good sequential throughput.
Some say that, in order to update (write into an already written stripe) on a RAID5, all blocks composing the stripe must be read in order to calculate the new 'parity'. This is false, the new parity is equal to "(replaced data) XOR (new data) XOR (existing parity)", therefore the logic only has to read the old block and the parity block, calculate then write the new blocks (data and parity).
Learn about Linux
Download Linux
Get Linux help
Re: Inspecting disk IO performance with fio
Posted by: Anonymous [ip: 192.168.0.139] on April 12, 2008 08:56 PMSome say that, in order to update (write into an already written stripe) on a RAID5, all blocks composing the stripe must be read in order to calculate the new 'parity'. This is false, the new parity is equal to "(replaced data) XOR (new data) XOR (existing parity)", therefore the logic only has to read the old block and the parity block, calculate then write the new blocks (data and parity).
Here are some hints and benches: http://www.makarevitch.org/rant/raid/
#