Should I ReInstall F14

[toddbailey]

Hi All,

I have a F14 x86_64 install built on a asus m/b w/ a single amd x64 dual core (1x 64x2 6000).
However, I've since transplanted the drive (500gb sata) to a HP workstation (xw8400) that is running dual xeon proc's w/dual core (2x Intel 5160).

The workstation runs as expected on the new box but just wondering if I should start over with a fresh install. I'd rather not as time spend updating and configuring is somewhat extensive.


Any gotya's in building a fedora system under one cpu and then moving to a new one?


but like I said every thing works as expected, system see's the 4 cores all the disks, video, audio, usb and other hardware with out any drama, etc. dmessg doesn't report any issues either.


thanks

[scott32746]

Hello,

I would say new hardware ( motherboard ) new install, but if you are working normal I would not do anything.
Replacing a harddrive, audio card is not a big deal, but motherboard is to me.

[jpollard]

You would only need to reinstall IF you change CPU architecture.

Nearly everything else is handled for you. As long as you are on an intel base CPU (whether from AMD or Intel) you should be good to go.

Only if you have custom video drivers would there be a possibility that things would go south.

[toddbailey]

I know it's a royal pain to perform a transplant in the windows world. Most of the time, I've always had to reformat and reinstall, but I have never had any problems with linux, such as having linux on a thumb drive, you can basically boot any machine you want. there is the problem of trying to boot a 64 bit os on a 32 bit cpu. You just don't get very far.
but the i386 appears to run on any intel/amd machine I've thrown at it.

But I just wanted to ask in case there were some unseen down stream issues that could appear in the future , and while one can swap a system drive across different machines.
Is it a recommended process?

[jpollard]

It is done fairly frequently when it comes to clusters. The boot server has disk images that are distributed to the disks of each node in the cluster.

Same disk image. I see no difference in that than in moving a disk from one node to another.

In the cluster it is done for speed - the first distribution gives each node a local disk to make subsequent boots from. This allows all nodes in the cluster to boot simultaneously without network contention. It can cut a boot time down from an hour to minutes. Yes, having multiple boot nodes helps - but there is still a delay. Note: updates may take a bit longer, but unless you are replacing the entire system it is still faster.

The only issues would be if the i386 (actually it is a 686 CPU - it won't boot on a real i386) is ever dropped.

Changing from 686 to x86-64 is the only thing that would come close to needing a re-install. Even then, it IS possible to build a 64 bit kernel and have an all 32 bit user space.