From owner-freebsd-hardware Sun Sep 9 23:18: 2 2001 Delivered-To: freebsd-hardware@freebsd.org Received: from zogbe.tasam.com (uta-ip196.ntc.off-campus.vt.edu [63.165.178.196]) by hub.freebsd.org (Postfix) with ESMTP id CF1E937B401 for ; Sun, 9 Sep 2001 23:17:56 -0700 (PDT) Received: from battleship (uta-ip196.ntc.off-campus.vt.edu [63.165.178.196]) by zogbe.tasam.com (8.11.6/8.11.6) with SMTP id f8A6Hdr66788; Mon, 10 Sep 2001 02:17:40 -0400 (EDT) Message-ID: <008901c139c0$4e1b7190$095f5f0a@battleship> From: "Joseph Gleason" To: "Kory Hamzeh" , "Brian Szymanski" Cc: References: <003701c13995$87b23cc0$14ce21c7@avatar.com> Subject: Re: EIDE Drive Errors - Please Help! Date: Mon, 10 Sep 2001 02:17:40 -0400 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.3018.1300 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.3018.1300 Sender: owner-freebsd-hardware@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org In my experience, the IBM drives with 15gb platters have had quality control problems. I would recommend against buying that generation if you can help it. Of 8 30gb IBM drives about 80% have failed. Of 2 75gb IBM drives, 100% have failed. (This is not a very big test set, I know) Never a complete failure...just an ever growing number of bad sectors...which as far as I can tell, IDE drives are not good at remapping. I don't know the interworkings of hard drives, so my statement there might be completely wrong. You can dump from one drive onto the other in one step. Something like: cd /mnt/new dump -au -0 -f - /dev/ad4s1a | restore -rf - Note: restore needs a /tmp directory that it can write some of its inode lists or something to. Make sure whatever contains /tmp is mounted R/W. ----- Original Message ----- From: "Kory Hamzeh" To: "Brian Szymanski" Cc: Sent: Sunday, September 09, 2001 21:11 Subject: RE: EIDE Drive Errors - Please Help! > You're scaring me, man! I've always have had very good luck with the > reliability of IBM hardware! Some kind person on the list pointed me to > software on IBM sites that allowed me to re-format and map the bad sectors > out. It seems to be fine right now and I don't have enough time right now to > swap drives and re-install the OS. The system has to be up by the morning. > > What is the easiest way of swapping both drives? I suspect I should do a > level 0 dump. Install the new drive. Install a minimum FBSD system and them > restore from tape on top of it?? > > Thanks, > Kory > > > > -----Original Message----- > > From: Brian Szymanski [mailto:brian.szymanski@cornell.edu] > > > > Ouch. I don't envy your situation. Those IBM Deskstar hard drives > > got rave > > reviews as being "the fastest EIDE hard drives ever," but apparently they > > tend to fall apart after a couple of months. I came close to buying one > > before by chance I happened to read some customer opinions of the product > > in which 75% of them said that it was "a great product if it only would > > f****ing work", see: > > > > http://computers.cnet.com/hardware/0-1092-418-1664463.html?tag=st.co.1092- > > 404-1664463.urev.1092-418-1664463 > > > > If at all possible, get a new hard drive and install on that ASAP. With > > EIDE drives, quality is pretty variable (but very good on certain > > products). I've had no trouble with my Maxtor, but YMMV. > > > > > I have a relatively new IBM DTLA-307045 45G UDMA/100 Hard disk drive > > > them seems to be having a lot of bad sectors. This is the slave drive > > > in the system. The first drive is a IBM DTLA-307030 and it is the > > > master. I'm using the proper shielded cable and the blue connector is > > > going to the motherboard EIDe controller. The slave drive is connected > > > to the middle connecter and the master is conncted to the end > > > connector. The motherboard is a ASUS CUSL2 with a 800MHZ P3 and 512MB > > > of RAM. The system is running RELEASE-4.3. > > > > > > This system was running NT 4.0 for several months without any problems, > > > although the slave drive was not used extensively. I have on partition, > > > ad1s1, which takes the entire drive. I used sysinstall to create the > > > slices, which I wanted to slices to match the master drive. So I did a > > > disklabel -e, and changed the slices to use ad1s1a, ad1s1e, and ad1s1f > > > (that is all I changed). I then did a "newfs -b 8192 -f 1024" on the > > > three filesystem. That all when fine. I mounted ad1s1f on /mnt can > > > copied about 20 meg, just as a test. I un-mounted it and ran fsck and > > > start seeing a lot of hard read error and "DMA problem -- using PIO" > > > errors. When I do a "dd if=/dev/ad1s1 of=/dev/null" I get the same > > > errors towards the end of the scan. > > > > > > I'm not that familiar with IDE drive -- I've been using SCSI mostly > > > until now. Do I need to get a new drive? Is there a way to map out the > > > bad sectors on this drive? > > > > > > I urgently need help -- I suppose to have did system up and running > > > online by Monday. > > > > > > Thanks, > > > Kory > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-hardware" in the body of the message > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hardware" in the body of the message