Gus Brown
02-07-2001, 11:34 AM
Arrrgh. Here is the scoop:
I need to replace two 30 GB drives in my Redhat 6.2 box with one 60 GB drive. I read many postings and 'how-tos' (obviously not enough of them, however) but am still coming up short. I currently have the following partitions and mount points:
hda1 DOS not mounted
hda5 /boot
hda6 /usr
hda7 /var
hda8 swap
hda9 /home
hda10 /
hdb1
hdb5 /data
hdc1 DOS not mounted
hdc5 /boottmp
hdc6 /usrtmp
hdc7 /vartmp
hdc8 swap
hdc9 /hometmp
hdc10 /tmp
hdc11 /www
hda = 30 GB drive, must go
hdb = 30 GB drive, must go
hdc = 60 GB drive, get (has) to stay
The first approach I tried involved mounting 'new' on hdc and then "cp -a /usr /new/" and so on. Then once everything was backed up I was to delete '/usr' using "rm -rf /usr" then "umount /dev/hda6 /usr" next "mkdir /usr" then "mount /dev/hdc6 /usr" followed by "cp -a /new/usr /usr" and so on until all the system directories had been transferred over. I got into big trouble with '/bin', '/sbin', '/etc' and a few others when those directories were not where the system thought they should be and I also kept getting errors saying that a given mount point was busy and that I could not umount it. Some reading revealed (from the MAN page, I believe) that if the system is using the directory I may get that message.
Q1: If I am getting this error because the system is using a given directory, how can this procedure for transferring partitions ever work?
Q2: How can I tell where the system directories that do not have their own partitions, like '/etc' and '/bin' reside?
Surely there must be some way of doing what I am trying to do but I am too dense to see it.
Q3: Assuming I successfully migrated everything so that it was just the way I wanted it, would I need to go in and relabel every occurence of 'hdc' as 'hda' since both 'hda' and 'hdb' will be getting jerked out of the box?
Thanks in advance for any light anyone is able to shed on this for me, sorry again for not being able to noodle this one all the way through...
Regards,
Gus
I need to replace two 30 GB drives in my Redhat 6.2 box with one 60 GB drive. I read many postings and 'how-tos' (obviously not enough of them, however) but am still coming up short. I currently have the following partitions and mount points:
hda1 DOS not mounted
hda5 /boot
hda6 /usr
hda7 /var
hda8 swap
hda9 /home
hda10 /
hdb1
hdb5 /data
hdc1 DOS not mounted
hdc5 /boottmp
hdc6 /usrtmp
hdc7 /vartmp
hdc8 swap
hdc9 /hometmp
hdc10 /tmp
hdc11 /www
hda = 30 GB drive, must go
hdb = 30 GB drive, must go
hdc = 60 GB drive, get (has) to stay
The first approach I tried involved mounting 'new' on hdc and then "cp -a /usr /new/" and so on. Then once everything was backed up I was to delete '/usr' using "rm -rf /usr" then "umount /dev/hda6 /usr" next "mkdir /usr" then "mount /dev/hdc6 /usr" followed by "cp -a /new/usr /usr" and so on until all the system directories had been transferred over. I got into big trouble with '/bin', '/sbin', '/etc' and a few others when those directories were not where the system thought they should be and I also kept getting errors saying that a given mount point was busy and that I could not umount it. Some reading revealed (from the MAN page, I believe) that if the system is using the directory I may get that message.
Q1: If I am getting this error because the system is using a given directory, how can this procedure for transferring partitions ever work?
Q2: How can I tell where the system directories that do not have their own partitions, like '/etc' and '/bin' reside?
Surely there must be some way of doing what I am trying to do but I am too dense to see it.
Q3: Assuming I successfully migrated everything so that it was just the way I wanted it, would I need to go in and relabel every occurence of 'hdc' as 'hda' since both 'hda' and 'hdb' will be getting jerked out of the box?
Thanks in advance for any light anyone is able to shed on this for me, sorry again for not being able to noodle this one all the way through...
Regards,
Gus