>T. Horsnell wrote: >> What's the situation with 'raw' and FC3? >> I want to get raw access to disks for test >> purposes but 'man raw' and 'whereis raw' >> show nothing. > >Same as for the rest of Linux: you can access a raw partition as >/dev/hdXY or /dev/sdXY (depending on technology), where X is a letter >and Y a number. So the first partition of your first IDE disk is >/dev/hda1. > >To refer to the whole disk, use /dev/hdX or /dev/sdX. > >Please note: get this wrong and you could delete your whole system. Be >careful. Use backups or a separate test system. > >Now the real question: what exactly are you trying to do, and why? >If you just want to see how fast your hard disk goes, try >hdparm -t /dev/hda >(or whatever). Hi James, Maybe I should have said direct I/O instead of raw. I discovered the O_DIRECT option to open(2) over the weekend and have been busy trying to use it. What I want to do (and what I have been doing on my Alpha system) is to thrash any new disk I plug in to my SCSI bus, before putting it to general use. I do this using a long test (at least 48 hours) of write/read/verify, of records full of random data, of random length, to random positions on the disk. I want to avoid the cache (which doesnt happen if I use /dev/sd?) hence the 'raw' requirement. And, joy of joys, I can now even use smartctl to monitor the disk health in the process. Its interesting to see the temperature rise as the disk starts to work hard... I hadnt heard of hdparm - what a useful tool. It shows buffered read rates on my 140GB 10Krpm SCSI drives of 75MByte/sec, but on my new 300GB 10Krpm one, I only get 65MBytes/sec. They're all Seagate - and all < 3 months old. Strange that. > >If you want to use raw copying, use dd (and be careful). > >Have I mentioned that you need to be careful? Yes, I'm a bit concerned about this as well. I need to build in some checks that I'm not doing this on a mounted filesystem. And maybe a pile of reads, just to flash the 'activity' light on the chosen drive, would help me avoid finger-troubles. Cheers, Terry. > >James. >-- >E-mail address: james | Whenever [Richard I] returned to England he always >@westexe.demon.co.uk | set out again immediately for the Mediterranean and > | was therefore known as Richard Gare de Lyon. > | -- '1066 and All That' > >-- >fedora-list mailing list >fedora-list@xxxxxxxxxx >To unsubscribe: http://www.redhat.com/mailman/listinfo/fedora-list >
