anyvody got a clue how the mkisofs.timestamp file is created??
From
http://mailman.linuxtag.org/pipermai...ry/001593.html
On Mon, Jan 13, 2003 at 02:52:07PM -0800, Rib Rdb wrote:
> A question I've often seen people asking is why their
> remastered knoppix cd's run slower than knoppix.
> People usually respond saying that Klaus uses a
> special ordering of the files in the KNOPPIX cloop
> file. When people ask how to generate the sort file,
> I've never seen a response. Am I just missing
> something? If so, where is the answer. If not, is
> there a reason that Klaus won't tell us how he does it?
There is no special reason, except that some readers may find it
confusing and obscure, but no magic is involved. The ordering
is still not as good as it should be, and I doubt that it makes
a significant difference. The proper way would be tracing all
open() calls in the kernel and gernerating the list from there,
but I found no convenient way to do this yet.
Currently I use the file access times as sorting criteria, which
may not always give the best results. Attached is an example script.
Your mileage may vary.
Regards
-Klaus KnopperNow, i'm interested in how well this works. I have seen Klaus write that no sorting at all may be just as fast sometimes. So i guess we can easily compate, and maybe improve this a little.Code:#!/bin/sh find / -anewer /KNOPPIX.build/mkisofs.timestamp -not \( -type d -o -path /KNOPPIX.build\* -o -path /tmp\* -o -path /proc/\* -o -path /mnt/\* \) -printf '%p %A@\n' | sort -rn +1 -t' ' | awk -F' ' '{print $1 " " ++i}' > /KNOPPIX.build/mkisofs.sort echo "/bin/mkdir 10090" >> /KNOPPIX.build/mkisofs.sort echo "/bin/chown 10080" >> /KNOPPIX.build/mkisofs.sort echo "/bin/cp 10070" >> /KNOPPIX.build/mkisofs.sort echo "/bin/ln 10060" >> /KNOPPIX.build/mkisofs.sort echo "/bin/rm 10050" >> /KNOPPIX.build/mkisofs.sort echo "/dev/* 20000" >> /KNOPPIX.build/mkisofs.sort echo "/etc/* 30000" >> /KNOPPIX.build/mkisofs.sort
anyvody got a clue how the mkisofs.timestamp file is created??
Heh, Klaus put an end to that wives tale, didnt he
I would also be interested in knowing how mkisofs.timestamp is generated.
I have a test remaster here that takes about 2 times longer to boot into KDE than the original ISO.
Even if there is a marginal speed improvement I would be happy.

OWC Mercury Elite Pro 2 Bay RAID Enclosure with USB 3.2 for Dual 3.5 inch HDDs
$87.99
EMC KTN-SL3 15-Bay Disk Array Enclosure w/2x 303-108-000E Cards, Dual PSU,No HDD
$239.99
Broadcom MegaRAID 9480-8e Fujitsu PRAID EP540e SAS RAID Card (Replac 9380-8e) 8i
$228.00
Broadcom 9600-24i RAID Controller PCIe Gen 4.1 Tri-Mode NEW 05-50111-01004
$649.99
Broadcom LSI 9305-16i 12Gbps SAS-3 PCIe x8 HBA 16-Ports Raid Controller
$64.99
MegaRAID 9580-8i8e 16-port 12G SAS/SATA/NVMe 8GB PCIe RAID - 05-50076-00
$1619.26
LSI 9305-16i SATA SAS 12Gbs Controller card PCIe 3.0 IT-Mode w/SAS SATA cable US
$120.73
NVME M.2 Adapter Card PCIE X16 Splitter Expansion Controller 4-Disk Soft RAID
$24.43
G-Technology G-SPEED Studio XL 48TB Thunderbolt 2 Storage - Housing Only
$200.00
Intel VROC Premium RAID Controller Card PCIe Add-On Module VROCPREMMOD
$265.00