Re: Horrendous RAIDframe reconstruction performance

[mlelstv%serpens.de@localhost (Michael van Elst)]

john%ziaspace.com@localhost (John Klos) writes:

>Next, raidctl doesn't handle NAME= for device yet:

>raidctl -v -a NAME=raid8tb1 raid0
>raidctl: ioctl (RAIDFRAME_ADD_HOT_SPARE) failed: No such file or directory

Yes, so far, only the config file knows about NAME= syntax.


>Finally, even though these are absolutely not SMR drives, rebuild is 
>running at 3.5 to 4 MB/sec, whether attached via USB-3 or directly 
>attached via SATA:

That's the reconstruction algorithm. It reads each stripe and if it
has a bad parity, the parity data gets rewritten.

This is very good for repairing a few missed writes e.g. after a power loss.

It is abysmally slow when a drive gets replaced as every stripe is read,
validated and then the parity is rewritten. If you don't have very smart
drives, that needs one rotation of the media per stripe.

For a raid mirror it is usually faster to copy the component manually
and then trigger a parity check. The check then just reads all disks
at high speed.

-- 
-- 
                                Michael van Elst
Internet: mlelstv%serpens.de@localhost
                                "A potential Snark may lurk in every tree."