I just got a maxtor 20GB HD. For some reason it's writing at normal speed(at least as far as I can tell) but read times are incredibly sloowww. I ran hdparm testing out -d0 -d1 -c0 -c1 and -c2 then tested w/ -t after each one. best was right after I ran -c1 I think but that was really slow... 64 MB in ~32 seconds... or about 1.8MB/s Not good... Any ideas? It was running fine under WinXP but I just couldn't stand the pain of running Windows anymore....

Are you using the 80 conductor cables?

Is it master or slave with a cdrom drive (slower)?

It's on an 80 conductor cable... but I've noticed no difference in the speed from when I had it on an older cable. Also It's set as master w/ a 3gb slave... the slave is really old but it is reading faster than the maxtor... the 3gb writes at like 1/3 the speed of the maxtor though. Oh, and the slave is a seagate if that makes any diff... and I just barely hooked the slave up like today and have noticed no difference in the speed of either drive

I was having a Maxtor drive and when it started slowering down, it died 1 weeks later, watch out!
Maxtor disk should be really fast, are you using the correct chipset in your kernel? Take a look at your cable too.

hmm... How do I tell if I'm using the correct chipset?
And I have looked at my cable... other drives work fine on it...

Just out of curiosity, is there any chance that a more optimized distro such as Gentoo would speed it up?

not really, this is all about the kernel, but maybe the big-slow-not-optimized redhat kernel does not choose the correct chipset. use "dmesg" to see the boot msg and try to locate a line like this one :
ALI15X3: IDE controller at PCI slot 00:04.0
ALI15X3: chipset revision 196
ALI15X3: not 100% native mode: will probe irqs later
ide0: BM-DMA at 0xd400-0xd407, BIOS settings: hdaMA, hdbMA
ide1: BM-DMA at 0xd408-0xd40f, BIOS settings: hdcMA, hddio
of course this msg is about MY chipset, try to find your and make sure it is really the one you should use... is your drive new btw? maybe it's gonna die, as I said.

Alright... Latest Update on my HD:
I didn't really need to use my computer last night so I made a couple of scripts and ran them:
script 1) ran hdparm -aXX /dev/hda using every other value starting at 64 working down to 0 and using hdparm -t /dev/hda 4 times after every hdparm -aXX /dev/hda
script 2) took the 3 XX values that had the best times and tested those ones each 8 times again.
script 3) hdparm -V
hdparm /dev/hda
hdparm -i /dev/hda
hdparm -I /dev/hda
hdparm -t /dev/hda (ran this line 8X)

script 1 was took an excuceating amount of time to execute because each hdparm -t command was averaging about 500kb to 1MB a second but I managed to grab 18, 40, and 42 as the best values to plug in for -a.
I have no idea how long script 2 took because I went to bed and just checked it this morning but I noticed that 18 had the best average of the three with only the first three test below 2MB/s while 40 had the best single time up at 3.5 MB/s
finally I just barely ran script 3 and it dropped below 3 MB/s ONCE! yeah... that's still crap but it's about 2x what I was getting before...
Oh, and if it makes any diff this was all done from runlevel 5 in a failsafe root session except #3 which I ran from a terminal in a root session of KDE (I needed to see the output of the whole script otherwise I would have run it outside of X... yeah I'm sure there's a real easy way to do that but I don't know it) anyhow here's the output of script 3:

And to Half_Elf, is this what you're looking for?

Alright... so how do I fix that little not in 100% native mode line?

VP_IDE: IDE controller at PCI slot 00:07.1
VP_IDE: chipset revision 6
VP_IDE: not 100% native mode: will probe irqs later

Well I suppose Multi-IO are less effective than real onboard chipset, but I remember you told us that your slave drive was faster... I think the only good thing you can do is to try your HD on another computer to see if there's some speed problem too. Maybe your multi-io card is too stupid to use the correct setting with your disk... or maybe the disk gonna die, I can't think about something else... very weird error

I/O is onboard... actually I've only got 3 cards in my box... vid card, NIC, and an old ISA 33.6k modem... onboard... it's on a VIA chipset umm... EPoX 8KTA3+ is the MoBo I think...

well it is strange, look like your kernel think your chipset is on a pci card there... I dunno what to tell you, really...

I know your pain. Our company has hundreds of IBM 300 PL desktops that came with the Maxtor 52049U4 20 GB Drives. Guess what. The drive stinks. It's not your computer system at all. Fortunatley our systems are under warranty, and IBM has had to replace probably over 100 of these drives. What I have been told from their reps is the controller board on the drive goes coo-koo and makes the drive retarded and slow. It works perfectly with the exception of the speed. Try to warranty your drive through Maxtor. I promise, your drive is bad. Run Maxtor Power Diags. If it takes longer than 2 hours to test your drive, you have your ammo for the Warranty Claim. Let them know it takes three days to run the Diags. That is about how long it has taken on some of our drives that we actually allowed the test to complete on. (We have some spare time). I relly hope this helps you out. I registered for this board to specifically reply to your post. I was searching google for a FW Flash for this drive and your post was the first hit. Later.

LOL so the drive is bad. Send it to Maxtor, they are nice with their warranty, you'll probably receive a 40GB if your 20GB is defective.