Partitions sharing cylinders

I partitioned my HDD with Yast on the 11.3 installation. Looking at the details in fdisk now, some of the partitions seem to be overlapping, i.e., use the same cylinders.


   Device Boot      Start         End      Blocks   Id  System
/dev/sdb1               1         249     1998848   82  Linux swap / Solaris
Partition 1 does not end on cylinder boundary.
/dev/sdb2   *         250       14594   115219457    5  Extended
/dev/sdb5             250        2208    15727635   83  Linux
/dev/sdb6            2739       **11133**    67424805   83  Linux
/dev/sdb7           **11133**       **11149**      121306   83  Linux
/dev/sdb8           **11149**       12453    10481664   83  Linux
/dev/sdb9           12454       14592    17180672   83  Linux

I hate to start over, but these include my /home and /boot partitions and could be critical. What’s more, there’s no guarantee that Yast wouldn’t make the same mistake a second time around. Is there a way to repair this with fdisk ot something else (hopefully w/o loss off data)? I want to partition a second disk, but I should take care of this first.

Well my first response to this is why so many partitions on one Hard Disk? I seldom suggest you have more than four per disk or more than two OS’ per disk. As to your question, in the case of logical partitions, you don’t have to stop at an entire cylinder boundary like is normal for primary partitions, but in the middle of the some 8K blocks that are present is OK. I don’t think there is an issue with what you are looking at here, though I don’t recommended such a setup. While large partitions waste size due to large allocation sizes, lots of small partitions waste space due to overhead and the tendency for the user to duplicate some common files on each partition. But this is just my opinion on the subject.

Thank You,

Beside swap, /, /boot and /home there is dual boot OS and an encrypted partition. I suppose one could be on the second disk, but that’s used for multimedia and needs as much space as possible.

As to your question, in the case of logical partitions, you don’t have to stop at an entire cylinder boundary like is normal for primary partitions, but in the middle of the some 8K blocks that are present is OK.

That’s good to know. :slight_smile: Thanks.

Your set-up is fine. The extended partition is a container to extend (just as the name suggests) beyond the normal 4 partition limit. When you use an extended partition, you are setting aside a portion of the overall hdd as a container. The logical drives you create with-in this container will share it’s address space with that of the container.
Thusly, If you have a 200GB extended partition and make 4 50GB logical drives, the extended partition will identify with the whole 200GB space and each logical will equate with 1-50GB, 51-100GB, 101-150GB, and 151-200GB respectively.

Use of more than 4 partitions is common place in UNIX/LINUX/BSD systems. I have 1 machine with 17 partitions on each of 3 internal drives, and 4 external drives. So this system actually uses a total of 119 partitions. Why I have done this is because of 2 MS OS’s, 3 Linux OS’s and many shared projects that need to be easily backed up and worked on from multiple OS’s where a fundamental lack of hardware support requires me to use specific OS’s with specific hardware but results need to be common to all.

On 2010-08-15 17:06, jdmcdaniel3 wrote:

> I don’t think there is an issue with
> what you are looking at here, though I don’t recommended such a setup.
> While large partitions waste size due to large allocation sizes, lots of
> small partitions waste space due to overhead

That was in FAT, IIRC. The allocation size does not vary as much in linux filesystems as it did in FAT.

What overhead? There is no overhead I know of. You can have as many partitions as you like. I have
disks with 20 or more.


Cheers / Saludos,

Carlos E. R.
(from 11.2 x86_64 “Emerald” GM (Elessar))