[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Problem with debian on Sparc

Hi Miroslav,

<miroslav.ludvik@centrum.cz> writes:

> Hi All,
> I instaled Debian on my old Sparc. In this Sparc was two 18 GB SCSI
> HDD in software mirror. I have two 73 GB HDD and I want change this
> disks in this server. This is production server and on this server are
> much data and installed aolication. It is not possible reinstall.

Ever heard about backups / rescuesystems?

> I change one HDD from 18 to 73 GB. I keep size partitions and one
> partition add (free space). Mirror is OK, but I have problem with
> silo, because when I edit silo and run silo, silo say me:

1. You don't need a reinstall. I for my part would add the old sdb into
   the machine and try booting from this. At the OBP  boot disk1
   Or if this wont work, boot d-i with rescue over net/by cdrom and then
   start a degraded array on the still working old sdb. if this works,
   resync the bootsector and partitiontable from sdb > sda, hotadd the
   missing raiddevices from sda. Now you should have your old system up.

> ==========================================================================================================
> 1) How make correct first hdd /dev/sda, which  I make problem my
> mistake?  This was sole boot device:(    I have fear from reboot. I
> think, when it is no problem, because on /dev/sda is silo and system
> and data will boot from md device 73 GB hdd (in degraded mode will be
> md device). But I am not sure.

where is your old /dev/sdb (18GB) couldn't you boot from this one? (if
you have an other sparc around use this one to test)

> 2)  What it is need make for change label and whole disk on /dev/sdb I need that sparc will be booted form boath HDD.and after this I change second 18 GB to 73.

If you have your system in proper state e.g. all raiddevices fixed, both
HD in sync, add your new 73GB as sdb. in fdisk choose auto. thereafter
you can partition the HD without reboot, and hotadd the missing
raiddevices. After sync, rerun silo. the remove 18gb sda and make sure
you can boot from sdb. then add new sda(73gb) and sync once again.

I would do all this with serial console attached and in single mode to
avoid that any production data will get corrupted.

If you really need somebody to do this job, contact me directly.


> Thank Mirek

Reply to: