> I have a related idea. If all of the information in a raidtab can be
> derived from a persistent superblock, there should be a tool that allows
> generating an actual raidtab file by reading a persistent superblock.
> This would be especially useful for dianostic procedures.
I have a perl script which includes an option to do this.
I run it to generate a `sample' raidtab file.
> Also, is there any downside to using persistent superblocks?
They use space at the end of the partition.
This means that you can't take an existing partition and make it into a
RAID1 just by running mkraid (but have to mkraid the new partition with a
`failed' disk, copy it over, add the original disk, etc).
It means that you cannot `extend' partitions.
> This is still an option in a raidtab, but perhaps it should be defaulted to
> enable and the capability of disabling persistent superblocks regarded as
> obsolete.
Agreed.
> On Sun, 2 Apr 2000, Michael T. Babcock wrote:
>> This of course suggests that the raidtab file is almost redundant itself and
>> could be replaced by command-line options to mkraid.
Agreed -- I did this some time ago for some `simple' cases.
I find that /etc/raidtab is `over used'.
1) it is needed by mkraid -- as above, it should not.
2) raidstart uses it -- should be possible to point it at one of the
constituent partitions (where it can find the PSB)
3) RedHat pokes around to do a `belt and braces' job.
