On Friday 13 May 2005 20:34, James Finnall wrote: > On Friday 13 May 2005 20:07, James Finnall wrote: > > On Friday 13 May 2005 14:34, [EMAIL PROTECTED] wrote: > > > Hi, <snip> > > I just thought of another possible solution I need to test. The write > operation using sdd with 32k byte aligned structures requires the use of > a raw device to bypass the VM system and file buffering issues. I > usually bind the drive to /dev/raw/raw1 for this. If I attempted to > verify using the same raw device to read the backup back for > verification it might not use the read-ahead that other devices use. I > will need to test this but it might be possible for the raw device to > bypass the kernel problem. > > James
Ok, I was able to verify the use of the raw device as success when the usual device fails doing backups. My notebook has had this verification failure on the last 5 backups. I modified the script to use /dev/raw/raw1 for the verification device and it passed without the I/O Error messsages. Then I attempted to verify manually using the normal device /dev/cdrom and it failed with all the I/O Errors at the very end of the verification process just like the last 5 backups. So it may be possible to work around the kernel read ahead issue on optical media if the drive is bound to a raw device and the raw device is used instead of the normal device that would go through the standard VM system. Instead of a sym link of /dev/cdrom pointing to /dev/hdd or /dev/scd0 it could point to /dev/raw/raw1 instead. That is of course if the raw device is bound to that drive. This could have other side effects like playing DVD's in xine for instance where you may well want the read ahead caching for performance. Of course the normal symlinks can remain and just create another if one would prefer like /dev/cdraw for instance. I have found with devfs these are rather easy to setup once the media is identified. But I am not sure devfs will work with creating raw devices as yet. I currently setup raw devices manually through the rc.local file on boot under 2.6.x and devfs. Hope it will be of some help to others. James -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
