Hello Per,
I do agree, since we start in english we should have go on the same way. Many non FR speaking people have more skills than I and other ideas than mine.
I though most of the part were already said here ... but let's summarize :
If ones matter is getting large storage devices you have different approach :
- use RAID5 array : you can aggregate many physical disk devices in a large logical one with redundancy mecanisms. This redundancy costs about one disk capacity : 100MB for a 3x100MB array
- use RAID0 array : you are stripping your data over different physical disk devices in a large logical one *
without* any redundancy
this is for the "physical" aspects of the problem : you 'll have to consider your newly create disk array as a single large device.
Then you have to consider the logical part of the things :
- you can use the old approach : create logical partition (primary and extended) on the new device and create the FS over these partitions,
- you can use the "new" (in SME world
) approach : create volume group and logical volumes then create you FS over this one.
The main advantage of LVM is that you can manage independantly you physical and logical volumes.
It's also a matter of which technology you are familiar with.
Once you have built your new data storage, you just have to join it to your system : look around at "add extra disk to sme".
et voilà
There are many way looking at this case. Mine is to keep it as simple and reliable as possible. Splitting storage area from "system one" sounds good for me.
I would also go for the LVM architecture for the storage, but it's morelikely because I am dealing with this technology very often than because it's a better choice.
I guess Chris, will summarize its results as soon he get something working
Gaston.
10 Replies
2471 Views