being as the man page recommends not to enable these on mounted file
systems i put it in /etc/rc just before it begins the mount processes.

Thanks for the help ;)

- james

On Tue, 2002-04-02 at 22:28, Wouter Slegers wrote:
> On Tue, Apr 02, 2002 at 03:42:43PM -0800, James A. Peltier wrote:
> > so when I issue the commands
> >
> > atactl /dev/wd0c readaheadenable
> > atactl /dev/wd0c writecacheenable
> >
> > these features should now be turned on.
> Probably, although strictly speaking you only _instructed_ the disk to
> enable these functions. It did not return an errormessage, so the disk
> did not complain, but there are drives out there that silently ignore
> certain commands. Usually this means that the command in question is
> irrelevant.
>
> > does it show up in identify if turned off ???
> "atactl /dev/wd0c identify" reports capabilities and command sets (such as
> enabling/disabling readahead, caching, SMART) under the headings "Device
> capabilities" and "Device supports the following command sets"
> respectively. This indicates the _support_ for these commands by the
> harddisk, not their status. In essence, a set of bitfields in the device
> are read that indicate whether the host can expect certain sets of
> commands to succeed, such as enabling SMART.
> Some of these capabilities can be selectively enabled/disabled, these
> show up under the heading "Device has enabled the following command
> sets/features". On modern drives, "read look-ahead" and "write cache"
> indicate that readaheadenable and writecacheenable succeeded. Older
> drives may ignore these commands and/or fail to report their status.
> In general most drives internally already use read look-ahead, so
> enabling read look-ahead will usually not increase performance.
>
> Note that you need to repeat these commands each time the device is
> reset, i.e. after a reboot. Putting the atactl-commands in /etc/rc.local
> is a good idea.
>
> With kind regards,
> Wouter
>
> --
> Wouter Slegers
> Your Creative Solutions
> "Security solutions you can trust and verify."
>
> [demime 0.98d removed an attachment 3K#Jype application/pgp-signature]

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]