Say, how do you specify the amount tested (ie. from 64 to 128?)

I think it depends on the program. Mandrake does one size and Gentoo does another. I didn't change it so I dunno.

Mine in Gentoo with 2.6 kernel.

Nothing special. Top one is a Western Digital 80GB and the bottom is a Maxtor 30GB. I do wish I could afford one of those SCSI thingys though. They seem to rock.

Later

:D :D :D

Here's Mine :)

I think from what everyone else Posted this is dang good!

-------------------------
root@pipewrench:~# hdparm -t /dev/hda

/dev/hda:
Timing buffered disk reads: 64 MB in 1.23 seconds = 52.03 MB/sec
root@pipewrench:~# hdparm -c 1 -d 1 -k 1 /dev/hda

/dev/hda:
setting 32-bit IO_support flag to 1
setting using_dma to 1 (on)
setting keep_settings to 1 (on)
IO_support = 1 (32-bit)
using_dma = 1 (on)
keepsettings = 1 (on)
root@pipewrench:~# hdparm -t /dev/hda

/dev/hda:
Timing buffered disk reads: 64 MB in 1.18 seconds = 54.24 MB/sec
root@pipewrench:~#
------------------------------

My Specs are:
Compaq Deskpro EN 733MHz
384MB RAM
40GB Maxtor DiamondMax ATA133 (8MB Cache)
Slackware Linux 9.1 <-----that's your boy!


Peace,
Whitehat

Timing buffer-cache reads: 648 MB in 2.01 seconds = 322.39 MB/sec
Timing buffered disk reads: 138 MB in 3.03 seconds = 45.54 MB/s

Maxtor 80GB 7200RPM

[root@localhost root]# hdparm -Tt /dev/sda

/dev/sda:
Timing buffer-cache reads: 1552 MB in 2.00 seconds = 776.00 MB/sec
Timing buffered disk reads: 214 MB in 3.01 seconds = 71.10 MB/sec
[root@localhost root]# hdparm -Tt /dev/sdb

/dev/sdb:
Timing buffer-cache reads: 1580 MB in 2.00 seconds = 790.00 MB/sec
Timing buffered disk reads: 158 MB in 3.00 seconds = 52.67 MB/sec

I thought SCSI was faster than that. :scratch: :scratch: :scratch:

Me :confused: :confused:

Later

:D :D :D :D

I don't get why you would assume a disk is naturally faster just because it uses the SCSI interface. Take a single hypothetical disk drive, for example. Now change the "connector" from whatever, to SCSI, to SATA, to IDE, to FC, to whatever. The only thing that has changed is the way the disk interfaces with its controller. The actual disk itself (platters, actuators, spindles, etc, etc) has remained exactly the same. So why would any particular disk be faster just because it uses a SCSI interface?

Besides, I think the scores for a couple of older disks are just fine.

I guess because they are so darn expensive. I always thought they were much faster. I guess they are just more reliable, maybe. :scratch: :scratch:

Oh well, learn something every day.

Later

:D :D :D :D

SCSI drives handle error corrections internally so they are more reliable. There are also options for hot swap replacements that keep highly available servers running when a drive is replaced - e.g., in a RAID environment.

Well, I am going to have to disagree that SCSI is more reliable. I have nearly 200 workstations at work and they are split about 50/50 between IDE and SCSI drives. The SCSI stations suffer drive failures about 3 times as often as the IDE stations. They are all about the same age (within 2-4 years old) and they are all several different brands. However Seagate drives seems to die more often than the rest. SCSI drives used to be faster because they supported higher spindle speeds and were a faster interface, but IDE drives are just as fast anymore. IMHO there really is no reason to buy a SCSI drive unless you are going to run 5 of them in a RAID config. RAID's are about the only places SCSI's shine.

I will agree that they are more expensive:)

It seems odd you are experiencing 3x more failures. If the mechanical parts were the same, I would expect the SCSI controllers to extend the life further than an IDE drive. We have almost all SCSI in about the same amounts, and some PC workstations with IDE. They SCSI drives and swapped around alot so I expect more failures (specially when dropped), but also realize if we did that with the IDE drives they would probably go much faster - connector life time as we use SCSI LVD connections.

The other place SCSI is good for is swapping drives in an out quick when you are bringing clusters up and down in software builds, and don't want to blow away a working set(s).

SCSI is more reliable than IDE, although the gap has lessend. Although there are numerous advantages to the SCSI interface, SCSI's biggest avantage is in the Enterprise, where the disks could be accessed by hundreds of users simultaneously, something a IDE disk would have nightmares about.

I agree. RAID's and hot-swappability are the only thing its got going for it. For a single user's workstation tho, in my opinion, IDE is the best choice. It's just as fast, more cost-effective, and in my experience more reliable. And yes, drive mirroring with IDE is a nightmare.....

hda:
Timing buffered disk reads: 80 MB in 3.02 seconds = 26.49 MB/sec

hdb:
Timing buffered disk reads: 154 MB in 3.01 seconds = 51.16 MB/sec

hda is 60 gigs, hdb is 120.
I'm not sure what brand/model hda is (probably some generic thing, came with my comp), but hdb is a Seagate.

# hdparm -Tt /dev/hda

/dev/hda:
Timing buffer-cache reads: 1436 MB in 2.00 seconds = 718.00 MB/sec
Timing buffered disk reads: 164 MB in 3.01 seconds = 54.49 MB/sec