|Rasputin||Apr 5, 2001 3:16 am|
|The Babbler||Apr 6, 2001 9:27 pm|
|Mike Meyer||Apr 7, 2001 2:48 am|
|Mike Barton||Apr 7, 2001 7:23 am|
|Dinesh Nair||Apr 7, 2001 8:49 am|
|Mike Meyer||Apr 7, 2001 11:52 am|
|Mike Meyer||Apr 7, 2001 10:00 pm|
|Gregory Bond||Apr 8, 2001 6:54 pm|
|Pete French||Apr 9, 2001 3:22 am|
|Rasputin||Apr 9, 2001 3:58 am|
|Mike Meyer||Apr 9, 2001 4:55 am|
|Pete French||Apr 9, 2001 5:00 am|
|Mike Meyer||Apr 9, 2001 5:05 am|
|Mike Barton||Apr 9, 2001 5:07 am|
|Pete French||Apr 9, 2001 5:32 am|
|Andrew Hesford||Apr 9, 2001 7:51 am|
|David Wolfskill||Apr 9, 2001 9:16 am|
|Andrew Hesford||Apr 9, 2001 10:38 am|
|Gerhard Sittig||Apr 9, 2001 11:51 am|
|Peter Jeremy||Apr 9, 2001 2:44 pm|
|John LoVerso||Apr 10, 2001 8:25 am|
|John LoVerso||Apr 10, 2001 8:28 am|
|Subject:||Re: Disklabel 101?|
|From:||Mike Meyer (mw...@mired.org)|
|Date:||Apr 9, 2001 4:55:28 am|
Pete French <pfre...@firstcallgroup.co.uk> types:
Mike Barton <mi...@dad.state.vt.us> types:
Are there any issues with placing swap first on the hard drive? Unless you insist on filling the drive, it seems to me that this swap arrangement would result in less stack travel.
Not that I know of. In fact, I'm pretty sure that one of my systems has root on s2, with swap on s1, which is earlier on the hard disk.
It doesnt work on the same disc however. Try and do an install of 4.2 creating a swap partition, then a root partition and it fails. Or always has in my expereince anyway. Its a bit of a nasty gotcha actually because it is not obvious what the problem is. I can see that it probably makes sense to have root frst then sawp (and they become a and b) so I dont do it that way anymore, but its still worth knowing about.
I'm doing it on one disk. Here's pstat -s for swap:
eve# pstat -s Device 1K-blocks Used Avail Capacity Type /dev/ad0s2b 135040 0 135040 0% Interleaved
So I'm swapping on s2. Here's what mount has to say about things:
eve# mount /dev/ad0s4a on / (ufs, local) devfs on /dev (devfs, local) /dev/ad0s4e on /usr (ufs, local, soft-updates) /dev/ad0s1 on /msdos (msdos, local) procfs on /proc (procfs, local)
So the Unix file systems are on s4. Here's the fdisk:
The data for partition 1 is: sysid 11,(DOS or Windows 95 with 32 bit FAT) start 63, size 7197057 (3514 Meg), flag 0 beg: cyl 0/ head 1/ sector 1; end: cyl 447/ head 254/ sector 63 The data for partition 2 is: sysid 165,(FreeBSD/NetBSD/386BSD) start 7197120, size 4192965 (2047 Meg), flag 0 beg: cyl 448/ head 0/ sector 1; end: cyl 708/ head 254/ sector 63 The data for partition 3 is: sysid 131,(Linux filesystem) start 11390085, size 4192965 (2047 Meg), flag 0 beg: cyl 709/ head 0/ sector 1; end: cyl 969/ head 254/ sector 63 The data for partition 4 is: sysid 165,(FreeBSD/NetBSD/386BSD) start 15583050, size 4401810 (2149 Meg), flag 80 (active) beg: cyl 970/ head 0/ sector 1; end: cyl 1023/ head 254/ sector 63
So slice 4 starts well after slice 2 on this disk. In fact, there's a Linux file system in between them.
I didn't use sysinstall, so that may refuse to do things like this. It may be that you tried to put root on something other than a, which causes the boot system to not work. I tried that on this system, then had to edit the disklabel to make the root partition a. Or it may be that -current does this but -stable doesn't, as this is a -current system.
The order of separate discs is not a problem. We have a Compaq server that insists on booting from da1 rather than da0, so I have / on da1, /usr on da0 and an interleved swap between them. Even in this case, however, I found it necessary to put the filesystems first on each disc, followed by the swap partitions.
I take it you mean root file systems? The default system layout puts root first, swap second, /var and then /usr, and has followed that pattern on pretty much every Unix I've run into - even systems with logical volume managers which add another layer between the disk and the file systems.
To Unsubscribe: send mail to majo...@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message