>   1. Is the  persistent-superblock 1 line required for each raid device,
>      or just for /boot?

It's not required for any raid device. "Persistent-superblock 1" (i.e. use
persistent superblocks) is the default, so you will get them unless you include
a "Persistent-superblock 0" line in your raidtab.

However, I'm not sure what you think the connection is with /boot. In the
traditional way of doing things, /boot is not on raid, so that would be
irrelevant. If you are going to have a shot at doing root raid using the
techniques suggested recently that don't need a separate /boot, then this would
also be irrelevant, as there would be little point in taking /boot off the root
partition in that case. Why do you need a separate raid device for /boot?

>   2. What is the purpose of the 1/0 after persistent-superblock?  Does
>      it just mean, "yes, use a persistent-superblock" or "no persistent
>      superblock for this raid device?  Is persistent-superblock 0 the
>      same as omitting a persistent-superblock statement?

1 means use persistent-superblock, 0 means don't use persistent-superblock.
Omitting the persistent-superblock line means that mkraid defaults to
"persistent-superblock 1", i.e. use persistent-superblock.

>   3. Does my raidtab file look okay?

> Here is my /etc/raidtab:
> raiddev /dev/md0
>         raid-level      1
>         nr-raid-disks   2
>         nr-spare-disks  0
>         chunk-size     4
>         persistent-superblock 1
>         device          /dev/hdc1
>         raid-disk       0
>         device          /dev/hdb1
>         raid-disk       1
          ^^^^^^^^^^^^^^^^^
          Lose this line

>         failed-disk     1

> raiddev /dev/md1
>         raid-level      1
>         nr-raid-disks   2
>         nr-spare-disks  0
>         chunk-size     4
> #        persistent-superblock 1
          ^^^^^^^^^^^^^^^^^^^^^^^^
          You will have persistent SB whether you uncomment this or not

>         device          /dev/hdc2
>         raid-disk       0
>         device          /dev/hdb2
>         raid-disk       1
          ^^^^^^^^^^^^^^^^^
          Lose this line

>         failed-disk     1

etc.
The various devices look OK apart from the "raid-disk 1" lines.

>   4. How do I determine whether the chunk-size is appropriate?

The appropriate chunk-size depends on the predominant file size on each device.
However, as you are using RAID-1, where all data is written to both disks, it is
not too significant. A default of "chunk-size 4" should be OK.

>   5. Do I need to set all of my partitions (hd[bc][1-3578]) to have
>      partition type 0xfd, or only my boot partition?

You want to set partitions to 0xfd on all devices which you want auto-recognized
at bootup. In your case, that looks like all of them.

>   6. Do I need to set the partition type of hdb[1-3578] to 0xfd before I
>      start up raid (with failed-disk), or just afterwards?

It doesn't matter. You will want to set it to 0xfd before you try to boot the
system from the raid devices. There were problems in the past with booting with
partitions of type 0xfd before the raid had been created. The raid code did not
free the inodes, so you could not run mkraid. I think this has been fixed, but
I'm not sure. It would be safest only to set them to 0xfd when you know that the
raids will be created before the next reboot.

Incidentally, I see you have left swap off raid. With the latest code, you can
safely put swap on RAID. If you want to prevent your system crashing when a disk
fails, you might want to do this.

Cheers,


Bruno Prior         [EMAIL PROTECTED]

Reply via email to