Has anyone tried the DiskOnChip life span calculator, or is it just a marketing gimmick? I need to upgrade a system to P4 and am trying to decide if I should go with DOC or CompactFlash; we've had problems with hard drives due to environmental issues. Unfortunately, customer insists on using Windows NT or Win2k and I'm not sure either one will work with DOC or CompactFlash.
If you use a CF to IDE adaptor then both NT and 2K should work.
The number I have heard tossed arround is "several days"; significantly less than a week,
That was for WinXP. I can't imagine any 'non WinCE' variants being significantly different.
SH7
Hello,
You need to use WinXp Embedded which has features to minimise the footprint and the writes to flash disc. If your customer insists on NT or 2K then as far as I am aware there is no MS support for prolonging the life of flash disc.
There is a free evaluation kit for XP embedded which allows you to build a system with a 100 day lifetime.
(All this from MS seminars - I have not actually done it !!)
Michael Kellett
OK, nobody else said it, so I will: this looks like you're in serious need of an upgrade to Customer XP --- the all-new "know what they're talking about" edition.
:-)
-- Hans-Bernhard Broeker (broeker@physik.rwth-aachen.de) Even if all the snow were burnt, ashes would remain.
[...re: using CF on a Win[2k|XP|NT] system ...] [...]
A basic feature of any OS of this type would let you configure a RAM disk and locate the swapfile and any temporary directories there. I know it can be done (i.e., have done it) for various flavors of UN*X. I can imagine that WinWhatever might not have such a feature, but would be somewhat surprised if that were the case.
Regards,
-=Dave
-- Change is inevitable, progress is not.
Perhaps I am missing something, but I thought there was a special file system for Flash which included a wear leveling mechanism to prevent constant writes to a single file from wearing out the flash drive. Is this not included in standard OS, only embedded OS?
I worked on a system that ran VxWorks and ran for ages using PCMCIA flash cards. We could poll each sector and see how much usage it had received. In testing we ran a utility to report the worst case usage and it indicated the drive would last for several years of continuous usage.
-- Rick "rickman" Collins rick.collins@XYarius.com Ignore the reply address. To email me use the above address with the XY removed. Arius - A Signal Processing Solutions Company Specializing in DSP and FPGA design URL http://www.arius.com 4 King Ave 301-682-7772 Voice Frederick, MD 21701-3110 301-682-7666 FAX
Maybe you could put some often accesed files on a ram disk, but a swap file is not one that would be useful. A swap file is used to hold data that has overflowed ram. Putting that in ram would just make the problem worse. Better using no swap file at all.
-- Rick "rickman" Collins rick.collins@XYarius.com Ignore the reply address. To email me use the above address with the XY removed. Arius - A Signal Processing Solutions Company Specializing in DSP and FPGA design URL http://www.arius.com 4 King Ave 301-682-7772 Voice Frederick, MD 21701-3110 301-682-7666 FAX
during
I know it doesn't make sense, but some systems work better swap files even if they're on a ram drive. I know in the past Windows really hated not having swap space even with a significant amount of ram.
I don't know about windows, but Linux can certainly run without a swap file. One thing that can reduce CF lifetime drastically in Linux is the fact that the filesystem wants to keep the file's *access* times up to date, i.e. even if you only read a file (e.g. when executing a binary) this causes a physical write access to the device. Fortuntely, this can be disabled by using the -noatime option when mounting the filesystem.
True.
Rob
-- Robert Kaiser email: rkaiser AT sysgo DOT de SYSGO AG http://www.elinos.com Klein-Winternheim / Germany http://www.sysgo.de
I cannot speak for Windoze, but creating swap space on a RAM based filesystem is NOT useful under Linux. As mentioned before, it just makes things worse.
Wolfgang Denk
-- Software Engineering: Embedded and Realtime Systems, Embedded Linux Phone: (+49)-8142-4596-87 Fax: (+49)-8142-4596-88 Web: www.denx.de Vor allem kein Gedanke! Nichts ist kompromittierender als ein Gedan- ke! - Friedrich Wilhelm Nietzsche _Der Fall Wagner_ (1888)
No. This is WRONG. A swap file in RAM will NOT improve performance. It will be much more efficient to use NO swap at all.
Wolfgang Denk
-- Software Engineering: Embedded and Realtime Systems, Embedded Linux Phone: (+49)-8142-4596-87 Fax: (+49)-8142-4596-88 Web: www.denx.de Any technology distinguishable from magic is insufficiently advanced.
...and on the number of power-cycles. It seems some CF cards perform wear-levelling only upon power-up.
Wolfgang Denk
-- Software Engineering: Embedded and Realtime Systems, Embedded Linux Phone: (+49)-8142-4596-87 Fax: (+49)-8142-4596-88 Web: www.denx.de Old programmers never die, they just branch to a new address.
Hello,
Embedded XP can have a large footprint, and push the RAM, and CPU requirements up. If there is no need for USB, or hot swap of PCMCIA, Embedded NT might be viable. It includes support for DiskOnChip, and one can disable the swap, IIRC.
M$ does not seem to market the NT product any more, even though it is sold under the Microsoft label. It can be obtained from Venturcom, or Avnet, though.
- Pat
This brings up another interesting topic: Compressed ram-based swapfile
A custom swapfile manager could lzh-compress the pages on swapout, and thus tradeoff between speed for often-used pages and small footprint for less-often-used pages.
Of course this will be dead slow when the memory system is at its limits, similar to a Windows system with 8MB and literally minutes of continous HD access.
However, it would provide an elegant and automatic solution to the problem of never-again-used-pages that you brought up!
Marc
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.