On Sat, 20 Nov 2010 00:10:02 +0100, Mick wrote about Re: [gentoo-user] migrating disks (from mounts to disklabels:
>On Friday 19 November 2010 19:19:34 David W Noon wrote: >> On Fri, 19 Nov 2010 17:00:04 +0100, Peter Humphrey wrote about Re: >> >> [gentoo-user] migrating disks (from mounts to disklabels: >> >On Friday 19 November 2010 14:42:23 Neil Bothwick wrote: >> >> On Fri, 19 Nov 2010 10:52:50 +0000, Mick wrote: >> >> > Also primary partitions which he does not seem to be using at >> >> > all have a slight edge over logical. >> >> >> >> Do you have any data on this? I generally use all logical >> >> partitions but could be persuaded to rethink. >> > >> >Well there must be one level of indirection on first access, since >> >the start of the logical partition has to be looked up in a >> >"primary" partition, but I can't imagine that being needed more >> >than once per reboot. >> >> Correct. >> >> The same applies to LVM2 or EVMS logical volumes: a small "lookup" >> penalty (a few milliseconds) when the filesystem is first >> activated/mounted, and as fast as the drive itself thereafter. > >Short of measuring the latency with some system (which I wouldn't know >how) I have experimented with setting the /boot partition on primary >and logical partitions and the difference (on a stopwatch) was >measurable in seconds betweeen having said partition on a primary and >having it on a logical. Furthermore, sda7 was slower than sda5. Unless you have the mother of all initrd's or initramfs's, you cannot have /boot on a logical partition -- only a primary partition, as BIOS interrupts will only access raw drives and primary partitions. If you do put /boot on a logical partition, you will pay the "lookup" overhead repeatedly as part of the early bootstrap process. Since you won't have a kernel running at that time. no caching, including device mapping, will be in force. It will be dog slow if /boot is not in the primary partition table. I always make my /boot partition /dev/sda1, which is the first primary on the first hard drive. >I haven't measured latencies for first mount and subsequent look ups. >I thought that it would be the same every time a partition fs is being >accessed, no? No. The absolute seek addresses of all partitions and logical volumes are cached by the kernel. Later accesses will always use the cached extent details. Resizing a logical volume rebuilds that part of the cache and, if you resize with the filesystem mounted, forces the filesystem driver to reread the cached extent information. -- Regards, Dave [RLU #314465] *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-* dwn...@ntlworld.com (David W Noon) *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
signature.asc
Description: PGP signature
