Which Hard Drive Manufacturer?
Stewart Buckingham
I've a long history of preferred drive manufacturers. Most recently I've
preferred Western Digital. Unfortunately, I've had 2 WD 120gb Caviar
drives, one gone south and the other going south, after less than 2
years of light home use. It's kinda tainted me against buying WD
replacement drives.
Is there any kind of consensus about which manufacturer or model
(ATA100/133) currently run reliably? I prefer to pay for quality rather
than cheapness (especially when it comes to drives).
Stu/2
Minnie Bannister
but the other machines here have mostly WD IDE drives. I had heard that
Seagate IDE drives were not so great, but they recently upped their
warranty to 5 years for all drives purchased since June 1, and I bought
a 160GB Seagate that I have not yet installed. Seagates are reputed to
be very quiet as well.
MB
On 09/01/04 06:46 am Stewart Buckingham put fingers to keyboard and
launched the following message into cyberspace:
Felix Miata
I've never bought and plan never to buy WD's designed for windoze drives
industry spec non-standard EIDE drives. OTOH, everything lately is
apparently equally bad. I've had to replace too many of everything. I've
been buying mostly Seagate since Maxtor swallowed Quantum and IBM
pitched their ball to Hitachi. Seagate is eerily quiet, and slower than
Hitachi on Sysbench.
--
"Blessed is the nation whose God is the Lord." Psalm 33:12 NIV
Team OS/2 ** Reg. Linux User #211409
Felix Miata *** http://members.ij.net/mrmazda/partitioningindex.html
-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 100,000 Newsgroups - 19 Different Servers! =-----
Paul Ratcliffe
wrote:
Having been all SCSI previously, I recently bought a pair of 160GB IDE
Seagates. They are blissfully quiet... compared to the horrendously loud
and squeaky IBM/Hitachi 36GB SCSI I bought before.
I wouldn't touch Hitachi IDE either as they seem similarly afflicted with
irritating squeak after seek.
The Seagates only have 1 yr. warranty so I expect they'll go phut shortly
after, but they are a damned sight cheaper than the equivalent SCSI (of
which I have not had a single failure in 9 years).
Trevor Hemsley
Ratcliffe <ab...@orac12.clara34.co56.uk78> wrote:
> The Seagates only have 1 yr. warranty
Not necessarily, they just changed them, some retrospectively, even if
it says 1 year on the tin.
http://www.pcworld.com/news/article/0,aid,117040,00.asp
--
Trevor Hemsley, Brighton, UK.
Trevor-...@dial.pipex.com
Stewart Buckingham
>>Is there any kind of consensus about which manufacturer or model
>>(ATA100/133) currently run reliably? I prefer to pay for quality rather
>>than cheapness (especially when it comes to drives).
>
>
> I am happy with my Hitachi Deskstar SATA drive, although I heard that
> Seagates would be a bit quieter (not that the Hitachi is very noisy).
> Seagate also recently upped they warranty, but I don't remember the new
> time period covered, but it is at least equal to Hitachi's now.
Does the SATA drive connect directly to ATA/EIDE connection?
Seagates were one of my preferred drives in the past. The old Seagates
were always noisy only to be outdone by the manufacturer which went out
of business about 3 years or so ago (name escapes me for the minute....
hangover).... SCSI drives.
Also surprising that the Hitachi's are a tad noisy since they were the
old IBM drives. I've never had any EIDE IBM drives, but the SCSI's were
rock solid and not particularly noisy.
> But I guess especially for disk drives you will always find people
> recommending brands that others do not agree with at all...
True. I've heard good and bad things about Hitachi drives.
Stu/2
Stewart Buckingham
>>I've a long history of preferred drive manufacturers. Most recently I've
>>preferred Western Digital. Unfortunately, I've had 2 WD 120gb Caviar
>>drives, one gone south and the other going south, after less than 2
>>years of light home use. It's kinda tainted me against buying WD
>>replacement drives.
>>
>>Is there any kind of consensus about which manufacturer or model
>>(ATA100/133) currently run reliably? I prefer to pay for quality rather
>>than cheapness (especially when it comes to drives).
>
>
> Having been all SCSI previously, I recently bought a pair of 160GB IDE
> Seagates. They are blissfully quiet... compared to the horrendously loud
> and squeaky IBM/Hitachi 36GB SCSI I bought before.
I used to be all SCSI also. I had 9gb & 18gb IBM's. I didn't think they
were particularly noisy. I'm glad I decided to switch to EIDE instead of
that noisy 36gb :)
> I wouldn't touch Hitachi IDE either as they seem similarly afflicted with
> irritating squeak after seek.
I'd heard that those IBM Deskstar EIDEs were dogs.
> The Seagates only have 1 yr. warranty so I expect they'll go phut shortly
> after, but they are a damned sight cheaper than the equivalent SCSI (of
> which I have not had a single failure in 9 years).
Minnie mentioned they've recently upped their warranty to 5 years for
all drives purchased since June 1st. This is either some kind of
confidence statement about their quality, or they don't expect to be
around in another five years. Warranties like that could, no will, kill
them if they have flakey drives.
I'm gonna check out their warranty, and because a few people have
mentioned how quiet they are, I'm gonna check out their range of EIDE
drives. Just cannot afford to go SCSI any longer. Every time I needed a
bigger SCSI drive, I had to buy a new controller also, due to changing
specs. No longer prepared to go thru that hoop.
Stu/2
Keith Cotroneo
Look for fluid drive bearing drives like seagate or maxtor. I can't
stand the wine of some
hard drives. The seagate FDB drives are very quiet. I have read that
maxtor fdb drives
are very quiet as well.
Keith
Scott
"Stewart Buckingham" <s...@despammed.com> wrote in message
news:2pnamrF...@uni-berlin.de...
> I used to be all SCSI also. I had 9gb & 18gb IBM's. I didn't think they
> were particularly noisy. I'm glad I decided to switch to EIDE instead of
> that noisy 36gb :)
are 4 or 5 of them
in my Intellistation at work and it's like sitting near a jet. I've taken to
wearing PlaneQuiet noise-cancelling
headphones in the office from time to time.
As for me, when I have to pay for drives, I'm buying Seagate drives lately
for the combination
of (apparent) reliability and the blissful quiet. You couldn't pay me to
put an IBM IDE drive in my machine
and, since Hitachi bought the plants, I guess that extends to Hitachi.
-Scott G.
Lutz
<cotroneore...@stnyremovenospam.stnyremovenospam.com> wrote:
>...
> The seagate FDB drives are very quiet. I have read that
> maxtor fdb drives
> are very quiet as well.
>
> Keith
>
SAMSUNG are very quiet too. Chose the ones with 8MB Buffer.
Regards
Lutz
--
--
email lutzPUNKTgeyerBEIalcanPUNKTcom
set BEI @ set PUNKT .
Mike Ruskai
No IDE drives are reliable. That's why they have warranties measuring
only 1 to 3 years.
I returned no fewer than 6 WD drives myself, and have since bought nothing
but IBM (now Hitachi Global Storage) SCSI drives. I had one go bad
shortly after buying it (it was probably bad from the start), which was
replaced under warranty. The replacement and all others (five total going
back several years) have worked perfectly since.
WD has a line of SATA drives (Raptor, I believe) which have SCSI-drive
specs, and also SCSI-drive prices and capacities. Otherwise, SATA is the
same as IDE, with regard to drive quality.
Use IDE/SATA if you want cheap capacity. Use SCSI if you want performance
and reliablity.
--
- Mike
Remove 'spambegone.net' and reverse to send e-mail.
Mike Ruskai
>Paul Ratcliffe wrote:
>
>> Having been all SCSI previously, I recently bought a pair of 160GB IDE
>> Seagates. They are blissfully quiet... compared to the horrendously loud
>> and squeaky IBM/Hitachi 36GB SCSI I bought before.
>
>I used to be all SCSI also. I had 9gb & 18gb IBM's. I didn't think they
>were particularly noisy. I'm glad I decided to switch to EIDE instead of
>that noisy 36gb :)
I have two IBM 36GB SCSI drives. Neither is noisy. Nor are any of the
18GB and 73GB drives.
I expect he got one with a bad bearing.
Minnie Bannister
> No IDE drives are reliable. That's why they have warranties measuring
> only 1 to 3 years.
Seagate recently increased the warranty on all its retail drives
(including IDE) to five years.
> I returned no fewer than 6 WD drives myself, and have since bought nothing
> but IBM (now Hitachi Global Storage) SCSI drives. I had one go bad
> shortly after buying it (it was probably bad from the start), which was
> replaced under warranty. The replacement and all others (five total going
> back several years) have worked perfectly since.
>
> WD has a line of SATA drives (Raptor, I believe) which have SCSI-drive
> specs, and also SCSI-drive prices and capacities. Otherwise, SATA is the
> same as IDE, with regard to drive quality.
>
> Use IDE/SATA if you want cheap capacity. Use SCSI if you want performance
> and reliablity.
Just to complicate matters, don't forget that the current (parallel)
SCSI is due to give way to Serial-Attached SCSI (SAS) "Real Soon Now": I
think I read that Seagate intends to bring its first ones to market in
4Q this year.
MB
Paul Ratcliffe
<spamten....@begonedynnaht.net> wrote:
>>> Having been all SCSI previously, I recently bought a pair of 160GB IDE
>>> Seagates. They are blissfully quiet... compared to the horrendously loud
>>> and squeaky IBM/Hitachi 36GB SCSI I bought before.
>>
>>I used to be all SCSI also. I had 9gb & 18gb IBM's. I didn't think they
>>were particularly noisy. I'm glad I decided to switch to EIDE instead of
>>that noisy 36gb :)
>
> I have two IBM 36GB SCSI drives. Neither is noisy. Nor are any of the
> 18GB and 73GB drives.
>
> I expect he got one with a bad bearing.
Bearing noise is a factor - all my SCSIs have been about the same for that
and much louder than the Seagate IDEs which I can barely hear above the
CPU and PSU fans.
On the IBM drives, there is an extra squeaking noise associated with head
seeking - even when nothing is happening, it does this damned thing every
minute where it writes performance or calibration data or something, which
causes a seek and a squeak.
It is not just my drive - as I said I have heard it on IBM IDEs as well.
Paul Ratcliffe
wrote:
> Hah! You must not have had the nasty IBM 10krpm u160 drives I have.
Yep, them's the ones.... IC35L036UWD210-0.
> There are 4 or 5 of them
> in my Intellistation at work and it's like sitting near a jet.
Heck, one's more than enough for me.
> As for me, when I have to pay for drives, I'm buying Seagate drives lately
> for the combination
> of (apparent) reliability and the blissful quiet. You couldn't pay me to
> put an IBM IDE drive in my machine
> and, since Hitachi bought the plants, I guess that extends to Hitachi.
Likewise.
Stewart Buckingham
> No IDE drives are reliable. That's why they have warranties measuring
> only 1 to 3 years.
Seagates are now 5 years.
> Use IDE/SATA if you want cheap capacity. Use SCSI if you want performance
> and reliablity.
I understand your sentiments exactly. I was all SCSI in the past, but I
couldn't cope any more with the high SCSI drive costs nor the need to
buy a new controller (due to technological advances) every time I needed
to increase capacity. I'm retired.
Eide also isn't very cheap if you have to buy a new drive every 2 years
because the drive goes kaput. Also, I tend to buy two at a time for
backup, normally retiring any old drives.
As an aside, I had a load of those 5 1/4" Quantum Bigfoots go south on
me in the past, but the 1 I have in my machine now justs keeps running
on and on. Now I said that, it'll probably die tomorrow ;)
Stu/2
Stewart Buckingham
> Just to complicate matters, don't forget that the current (parallel)
> SCSI is due to give way to Serial-Attached SCSI (SAS) "Real Soon Now": I
> think I read that Seagate intends to bring its first ones to market in
> 4Q this year.
One of the reasons I stopped buying SCSI drives. Changing specs required
new controller :(
Stu/2
tho...@antispam.ham
> Minnie Bannister wrote:
My understanding is that SCSI has always been backward compatible, so
you could plug a newer drive into an older controller card and have it
work properly, provided, of course, that the connectors had the same
number of pins. Obviously the controller wouldn't be able to take
advantage of the capabilities of the newer drive.
John Thompson
> Mike Ruskai wrote:
>
>> No IDE drives are reliable. That's why they have warranties measuring
>> only 1 to 3 years.
>
> Seagates are now 5 years.
Oh-oh. Should I start worrying?
Device: SEAGATE SX4464524 Version: B410
Serial number: LF113956
Device type: disk
Local Time is: Sun Sep 5 10:09:40 2004 CDT
Device supports SMART and is Enabled
Temperature Warning Disabled or Not Supported
SMART Health Status: OK
Vendor (Seagate) cache information
Blocks sent to initiator = 2681538369
Blocks received from initiator = 9713496
Blocks read from cache and sent to initiator = 34582242
Number of read and write commands whose size <= segment size = 892516
Number of read and write commands whose size > segment size = 0
Vendor (Seagate) factory information
number of hours powered up = 46877.57
46877 hours = ~5.35 years...
--
-John (jo...@os2.dhs.org)
Stewart Buckingham
>>Seagates are now 5 years.
> Oh-oh. Should I start worrying?
That was my sentiments exactly. Either their drives are now that good or
they(/we?) will all go out with a big bang. Ahem!
> Vendor (Seagate) cache information
> Blocks sent to initiator = 2681538369
> Blocks received from initiator = 9713496
> Blocks read from cache and sent to initiator = 34582242
> Number of read and write commands whose size <= segment size = 892516
> Number of read and write commands whose size > segment size = 0
> Vendor (Seagate) factory information
> number of hours powered up = 46877.57
Blimey! They keep all that info on the drive nowadays? How do you access
that information?
> 46877 hours = ~5.35 years...
Definitely scheduled for "retirement" real soon now ;)
Stu/2
Wolfi
> Eide also isn't very cheap if you have to buy a new drive every 2 years
> because the drive goes kaput.
Just a short correction for possible future use:
"kaput" is written correctly like this: "kaputt" ;-)
Wolfi
Minnie Bannister
One of my 5-yr-warranty Seagate SCSI drives is still going strong after
7 years. My other Seagates are all newer. Another computer has a
1-yr-warranty Maxtor that is still going strong after 4 years or more,
and our son's computer has a 1-yr-warranty WD that's been running for
two or three years.
But, being a belt-and-suspenders type, I do full backups every weekend
and differential backups every night -- all to tape.
If it makes a difference (it probably does), our machines seldom get
turned off except to install new hardware.
MB
On 09/06/04 05:12 am Stewart Buckingham put fingers to keyboard and
launched the following message into cyberspace:
>>> Seagates are now 5 years.
Stewart Buckingham
My Chambers Twentieth Century Dictionary allows both spellings. Perhaps
I should move into the twentyfirst century :)
Stu/2
Skylar Thompson
> John Thompson wrote:
>
>> 46877 hours = ~5.35 years...
>
> Definitely scheduled for "retirement" real soon now ;)
Don't be so sure. Over the summer, I dragged out my first computers. It's
an Epson QX-16 with a 20MB Seagate MFM drive. It booted up flawlessly, even
though it's around 20 years old. Seagate knows how to make drives.
--
-- Skylar Thompson (sky...@os2.dhs.org)
-- http://www.os2.dhs.org/~skylar/
Mike Ruskai
Aside from some oddball non-compliant devices, that's completely true.
You can plug a SCSI-1 device into a U-320 LVD/SE controller, and it will
work. You can plug a U-320 LVD/SE device into a SCSI-1 controller, and it
will work (given an ID from 0 to 7, of course).
To get around the issue of pin counts, you simply need an adapter. With
SAS, the physical interface will change, but because it's still SCSI, an
adapter will allow an SAS drive to work with an existing parallel SCSI
controller.
Which makes one wonder where this non-issue of changing specs requiring
new hardware comes from.
Especially since precisely the same is true of IDE - to use each newer IDE
interface (ATA66/100/133, etc.), you need a new IDE controller. But the
newer drives still worked with existing IDE controllers, at the slower
interface speed. Of course, with SATA, largely considered the newest
"IDE" (it's closer to SCSI than IDE, though), you need a whole new
controller, period.
The one and only legitimate argument against SCSI is price. And the
simple rebuttal to that one valid argument is that old adage, "You get
what you pay for."
John Thompson
> John Thompson wrote:
>
>> Vendor (Seagate) cache information
>> Blocks sent to initiator = 2681538369
>> Blocks received from initiator = 9713496
>> Blocks read from cache and sent to initiator = 34582242
>> Number of read and write commands whose size <= segment size = 892516
>> Number of read and write commands whose size > segment size = 0
>> Vendor (Seagate) factory information
>> number of hours powered up = 46877.57
>
> Blimey! They keep all that info on the drive nowadays? How do you access
> that information?
If you drive is SMART compliant (most are, these days) you can use
"smartctl" to access the info:
http://os2power.dnsalias.com/yuri/software/smartctl-b1.zip
--
-John (jo...@os2.dhs.org)
Stewart Buckingham
>>Blimey! They keep all that info on the drive nowadays? How do you access
>>that information?
> If you drive is SMART compliant (most are, these days) you can use
> "smartctl" to access the info:
>
> http://os2power.dnsalias.com/yuri/software/smartctl-b1.zip
OK. I downloaded this but when trying to run it informs me that it
cannot find LIBC03. Do you have a url for that?
Stu/2
John Thompson
IIRC, that's part of the emx runtime package. You should be able to get
it from hobbes.
--
-John (jo...@os2.dhs.org)
William L. Hartzell
It is part of Innotek Gcc distro (not current). I've got that dll that
came with Warpvision that is from Netlabs.org site.
--
Bill
Thanks a Million!
Britt Turnbull
I found it as a dload in Chuck's eCSMT package......
I have SCSI drives here and I can't for the life of me get it to work,
it just reports no ATAPI drives found.... any clues anyone....
thanks,
--
Britt - britt os2ecs co uk...
Winxx.. keep the apps.. junk the OS.. => http://www.ecomstation.com
System uptime is 15 days 11:58 hrs/mins
Stewart Buckingham
>>>>Blimey! They keep all that info on the drive nowadays? How do you access
>>>>that information?
>>
>>>If you drive is SMART compliant (most are, these days) you can use
>>>"smartctl" to access the info:
>>>
>>>http://os2power.dnsalias.com/yuri/software/smartctl-b1.zip
>>
>>
>>OK. I downloaded this but when trying to run it informs me that it
>>cannot find LIBC03. Do you have a url for that?
>
>
> IIRC, that's part of the emx runtime package. You should be able to get
> it from hobbes.
I have EMX installed, but that file is called EMXLIBC03.DLL
Stu/2
Stewart Buckingham
>> IIRC, that's part of the emx runtime package. You should be able to
>> get it from hobbes.
>>
> It is part of Innotek Gcc distro (not current). I've got that dll that
> came with Warpvision that is from Netlabs.org site.
Can you send me the dll? I don't really want to install the warpvision
.wpi just to get it.
Stu/2
Stewart Buckingham
>>>If you drive is SMART compliant (most are, these days) you can use
>>>"smartctl" to access the info:
>>>
>>>http://os2power.dnsalias.com/yuri/software/smartctl-b1.zip
>>
>>
>>OK. I downloaded this but when trying to run it informs me that it
>>cannot find LIBC03. Do you have a url for that?
> I found it as a dload in Chuck's eCSMT package......
>
> I have SCSI drives here and I can't for the life of me get it to work,
> it just reports no ATAPI drives found.... any clues anyone....
I checked and checked and eventually found it in my ecsMT "old
downloads" directory. Copied it over. Now Smartctl works, but I could
figure out how to specify the device (all the examples were for linux).
Oh well, time to RTFM.
Stu/2
Alex Taylor
<spamten....@begonedynnaht.net> wrote:
> The one and only legitimate argument against SCSI is price. And the
> simple rebuttal to that one valid argument is that old adage, "You get
> what you pay for."
I've been all-SCSI for several years. I have no complaints about the
reliability, but the cost/size ratio is just ruinous. Ever since I went
all-SCSI, I've been in a more or less constant state of being desparately
short of disk space. I don't think I can put up with it much longer.
I mean, I look at the price catalogues and see that the cost of a 64 GB
SCSI drive will get me 300 GB IDE, and think 'why am I doing this to
myself?'
I originally went to SCSI for two reasons:
(1) I wanted the most kick-ass drive performance I could get at the
time, given that I wanted to do program development and CD-writing.
(To the SCSI setup I added HPFS386 with a 128 MB cache.)
(2) I needed at least two hard drives for various reasons, but I already
had an IDE LS-120 drive, and I also wanted separate DVD-ROM and CD-RW
drives. That would have been five IDE devices. I thought SCSI
was a better solution; and hey, if I was going SCSI, why not go
-all- SCSI?
As usual, though, I underestimated the rate at which disk consumption
would grow. Virtual PC just EATS up disk space. So do mp3s, so do
digisubs, so do CVS sandboxes...
I'm seriously considering going SATA in my next PC (which I'll probably be
getting within the next 6-10 months). I'll probably keep my SCSI DVD-ROM
drive, though, since it's one of the last models with no hardware region
locking.
--
Alex Taylor
http://www.cs-club.org/~alex
John Thompson
> I have SCSI drives here and I can't for the life of me get it to work,
> it just reports no ATAPI drives found.... any clues anyone....
Try using the "-d SCSI" switch to tell it to look for SCSI devices.
--
-John (jo...@os2.dhs.org)
Stewart Buckingham
>>I have SCSI drives here and I can't for the life of me get it to work,
>>it just reports no ATAPI drives found.... any clues anyone....
>
>
> Try using the "-d SCSI" switch to tell it to look for SCSI devices.
I tried "smartctl -d ata" but it says it needs a device name as the last
argument. I tried 0, 1, c:. None worked. What is my device name likely
to be?
Stu/2
tho...@antispam.ham
> Virtual PC just EATS up disk space.
That's something I've wondered about. When I read through the
documentation for Virtual PC, it mentioned how the virtual
file system is contain in a single native file. Now, if that
single native file is on HPFS, which has a 2 GB size limit,
does that mean your virtual machine is limited to a 2 GB
partition size?
As for the SCSI vs. IDE decision, have you investigated using
one of those IDE to SCSI convertors? Let's you buy the less
expensive high-capacity IDE drives while attaching them to
your SCSI adapter.
William L. Hartzell
Sent to the address in your header.
James J. Weinkam
Not according to Webster's Third International, which lists kaput first and
"also kaputt" as an afterthought.
William L. Hartzell
I think that is language/country Dependant. Webster is an American
dictionary.
Stewart Buckingham
> As for the SCSI vs. IDE decision, have you investigated using
> one of those IDE to SCSI convertors? Let's you buy the less
> expensive high-capacity IDE drives while attaching them to
> your SCSI adapter.
Why would you want to do that? Where can you get such converters?
Stu/2
James J. Weinkam
So, what's wrong with that?
Try Google searches on kaput and kaputt.
William L. Hartzell
James J. Weinkam wrote:
<snip>
>
> So, what's wrong with that?
>
> Try Google searches on kaput and kaputt.
German, Wolfi is German. With one t is an American spelling dating from
before WW I. Google affirms the above.
Dave Saville
>
>
>On Mon, 06 Sep 2004 08:40:26 GMT, Mike Ruskai
><spamten....@begonedynnaht.net> wrote:
>> The one and only legitimate argument against SCSI is price. And the
>> simple rebuttal to that one valid argument is that old adage, "You get
>> what you pay for."
>
>I've been all-SCSI for several years. I have no complaints about the
>reliability, but the cost/size ratio is just ruinous. Ever since I went
>all-SCSI, I've been in a more or less constant state of being desparately
>short of disk space. I don't think I can put up with it much longer.
>
>I mean, I look at the price catalogues and see that the cost of a 64 GB
>SCSI drive will get me 300 GB IDE, and think 'why am I doing this to
>myself?'
So have I - but a good source of, sometime used, SCSI kit is your local
SUN dealer - find one who deals in used kit.
For example here in the UK I can get one of those cabinets to hold a
dozen SCA discs for about 70GBP and fill it with small cheap SCSI
drives, which nobody wants anymore. These are often unused as someone
orders a SUN box which say comes with a 32 gig drive but they want
something bigger, so the dealer takes out the 32, stuffs in what the
customer wanted and sells the 32 off cheap - 'cos no one else wants it
either.
HTH
--
Regards
Dave Saville
NB Remove no-spam- for good email address
Alex Taylor
wrote:
>>I've been all-SCSI for several years. I have no complaints about the
>>reliability, but the cost/size ratio is just ruinous. Ever since I went
>>all-SCSI, I've been in a more or less constant state of being desparately
>>short of disk space. I don't think I can put up with it much longer.
>>
>>I mean, I look at the price catalogues and see that the cost of a 64 GB
>>SCSI drive will get me 300 GB IDE, and think 'why am I doing this to
>>myself?'
>
> So have I - but a good source of, sometime used, SCSI kit is your local
> SUN dealer - find one who deals in used kit.
>
> For example here in the UK I can get one of those cabinets to hold a
> dozen SCA discs for about 70GBP and fill it with small cheap SCSI
> drives, which nobody wants anymore. These are often unused as someone
> orders a SUN box which say comes with a 32 gig drive but they want
> something bigger, so the dealer takes out the 32, stuffs in what the
> customer wanted and sells the 32 off cheap - 'cos no one else wants it
> either.
I wouldn't buy a used hard drive for my production box, same way I
wouldn't buy a used monitor.
I will point out that in all my years of computing (since about 1990), I
have only ever had one hard drive die on me. It was a used Micropolis
SCSI drive, which proved flaky from the start, and died within a year.
In any case, I don't know of any local Sun dealers.
Alex Taylor
<tho...@antispam.ham> wrote:
> That's something I've wondered about. When I read through the
> documentation for Virtual PC, it mentioned how the virtual
> file system is contain in a single native file. Now, if that
> single native file is on HPFS, which has a 2 GB size limit,
> does that mean your virtual machine is limited to a 2 GB
> partition size?
The virtual container file seems to grow dynamically, even past 2 GB if
necessary. IIRC, it creates a second file to handle the overflow past 2
GB. (And presumably a third, if needed.) The virtual guest sees them as
a single partition.
tho...@antispam.ham
>> As for the SCSI vs. IDE decision, have you investigated using
>> one of those IDE to SCSI convertors? Let's you buy the less
>> expensive high-capacity IDE drives while attaching them to
>> your SCSI adapter.
> Why would you want to do that?
To get a 200 GB "SCSI" drive for much less than the cost of a
real 146 GB SCSI drive.
> Where can you get such converters?
Directly from the manufacturer. Addonics makes them, for example.
There are others.
Peter Higson
> Stewart Buckingham wrote:
>
> > Eide also isn't very cheap if you have to buy a new drive every 2 years
> > because the drive goes kaput.
>
> Just a short correction for possible future use:
> "kaput" is written correctly like this: "kaputt" ;-)
>
> Wolfi
In English it is spelled kaput, in German kaputt, from the French
capot. Etre capot is to be without tricks in piquet etc (i.e. ruined,
done for).
Source: Concise Oxford Dictionary 9th edition.
Hope that helps to clarify.
Peter
--
Peter Higson
Victoria, BC, Canada
Christian Hennecke
wrote:
> I've a long history of preferred drive manufacturers. Most recently I've
> preferred Western Digital. Unfortunately, I've had 2 WD 120gb Caviar
> drives, one gone south and the other going south, after less than 2
> years of light home use. It's kinda tainted me against buying WD
> replacement drives.
>
> Is there any kind of consensus about which manufacturer or model
> (ATA100/133) currently run reliably? I prefer to pay for quality rather
> than cheapness (especially when it comes to drives).
I'd go for Samsung. Very fast and silent at the same time.
--
"I smell blood and an era of prominent madmen." - W.H. Auden
Stewart Buckingham
> Stewart Buckingham writes:
>
>
>>>As for the SCSI vs. IDE decision, have you investigated using
>>>one of those IDE to SCSI convertors? Let's you buy the less
>>>expensive high-capacity IDE drives while attaching them to
>>>your SCSI adapter.
>
>
>>Why would you want to do that?
>
>
> To get a 200 GB "SCSI" drive for much less than the cost of a
> real 146 GB SCSI drive.
Besides cost, are there any other (technical) advantages?
Stu/2
John Thompson
> Stewart Buckingham writes:
>
>>> As for the SCSI vs. IDE decision, have you investigated using
>>> one of those IDE to SCSI convertors? Let's you buy the less
>>> expensive high-capacity IDE drives while attaching them to
>>> your SCSI adapter.
>
>> Why would you want to do that?
>
> To get a 200 GB "SCSI" drive for much less than the cost of a
> real 146 GB SCSI drive.
I did it by stringing together a bunch of used 47G SCSI drives into a RAID
array...
--
-John (jo...@os2.dhs.org)
tho...@antispam.ham
Not that I know of. If there was a technical advantage, one would
think that it would have been implemented in real SCSI drives.
tho...@antispam.ham
>> Stewart Buckingham writes:
>>>> As for the SCSI vs. IDE decision, have you investigated using
>>>> one of those IDE to SCSI convertors? Let's you buy the less
>>>> expensive high-capacity IDE drives while attaching them to
>>>> your SCSI adapter.
>>> Why would you want to do that?
>> To get a 200 GB "SCSI" drive for much less than the cost of a
>> real 146 GB SCSI drive.
> I did it by stringing together a bunch of used 47G SCSI drives into a RAID
> array...
Now that's an uncommon size. I've seen SCSI drives of 4.5 GB, 9 GB,
18 GB, 36 GB, 73 GB, and now 146 GB. Occasionally I see the odd size
like yours, but not very often.
There was a time when disk drive capacities were the same between
the IDE and SCSI. Lately, however, IDE drives have been 80 GB,
120 GB, 160 GB, and 200 GB. I wonder why SCSI drives don't have
the same capacities but just different interfaces?
John Thompson
> John Thompson writes:
>
>>> Stewart Buckingham writes:
>
>>> To get a 200 GB "SCSI" drive for much less than the cost of a
>>> real 146 GB SCSI drive.
>
>> I did it by stringing together a bunch of used 47G SCSI drives into a RAID
>> array...
> Now that's an uncommon size. I've seen SCSI drives of 4.5 GB, 9 GB,
> 18 GB, 36 GB, 73 GB, and now 146 GB. Occasionally I see the odd size
> like yours, but not very often.
http://www.seagate.com/support/disc/specs/scsi/st446452w.html
I have 5 of these in the RAID array.
--
-John (jo...@os2.dhs.org)
Stewart Buckingham
OK. I won't run out and buy one of those $70 adapters then to use with
my Adaptec 2940UW I have hanging around here somewhere, collecting dust.
Err... I just thought... you answered the question "technical advantages
over real scsi". I meant "technical advantages over real atapi".
Stu/2
tho...@antispam.ham
Ah, that's easy. You can hang 15 SCSI devices off of one adapter,
unlike ATAPI. The Addonics adapter claims to support the Ultra 160
SCSI standard, which theoretically offers higher throughput than
ATA100, ATA133, or even SATA 150. In practice you'll be limited
by what a single drive can deliver to the IDE-SCSI interface.
Stewart Buckingham
OK. Just for fun...
15 Seagate 200GB ATA100 Drives @ $120 = $1800
15 ATA/SCSI Ultra 160 adapters @ $70 = $1050
1 Adaptec 29160 Ultra 160 controller = $250
1 Internal to External Cable convert = $20
2 8 Drive enclosures/cables @ $280 = $560
Thats $3680 plus shipping, say $4000 for one kick-ass drive system :)
Stu/2
Melvin Klassen
<jo...@starfleet.os2.dhs.org> wrote:
> > Blimey! They keep all that info on the drive nowadays? How do you access
> > that information?
>
> If you drive is SMART compliant (most are, these days) you can use
> "smartctl" to access the info:
>
> http://os2power.dnsalias.com/yuri/software/smartctl-b1.zip
A simpler program:
OS/2 Warp 4 -> Warpcenter -> OS/2 System -> Problem Determination
Tools -> Hard File Monitor.
with much less output.
Minnie Bannister
compatible?
I am sure I had one many years ago but don't remember any more.
MB
On 09/10/04 10:48 am Melvin Klassen put fingers to keyboard and launched
the following message into cyberspace:
Mike Ruskai
SCSI supports more devices. IDE has a max of two per channel. You can
string 15 IDE drives on a 16-bit SCSI chain with 15 of those devices.
--
- Mike
Remove 'spambegone.net' and reverse to send e-mail.
Daniela Engert
>>Besides cost, are there any other (technical) advantages?
>
>
> SCSI supports more devices. IDE has a max of two per channel. You can
> string 15 IDE drives on a 16-bit SCSI chain with 15 of those devices.
>
Which happens to be the very reason why both of them are dead-enders.
Ciao,
Dani
tho...@antispam.ham
> Mike Ruskai wrote:
Please explain. Do you have a better alternative?
Daniela Engert
Fast parallel interfaces (P-ATA, P-SCSI) are pleagued by the same
serious deficits:
1) high number of connector pins
a) reduces reliability
b) increases variance of electrical conditions
2) the "speed wall":
all of the signalling lines need to reach stable values within a common
sampling time "window". Faster transfer speeds narrow the window, while
point 1b) above increases the probability of missing the window - even
more so with longer cables and more units attached to the cable. Dealing
with this problem increases system costs exponentially.
3) compatibility with 1st generation hardware:
the protocols have parts which need to be understood by all devices and
need thusly to be carried out in the common denominator transfer mode -
usually the slowest one. In the fastest implementations, the "1st
generation mode" parts (arbitration, command delivery) eat up up to the
half of the bandwidth available (worst case).
4) buses: attaching multiple units
a) increases problems 1b) and thus 2)
b) potentially reduces the amount of bandwidth available to individual
units
c) requires arbitration of the shared medium,
d) causes collitions, thereby increasing latency
The solution is serial interfaces like S-ATA and SAS:
1) as few pins as possible
2) the "speed wall" is much less paramount
3) drop support for legacy devices (or use bridges)
4) allow point-to-point connections only, with additional switching
devices if required (f.e. SATA/SAS port multipliers).
Have a look at the imminent future: PCI-X/PCI-e SATA/SAS (even
bilingual) host adapters with 4/8 (or more) ports running at 3 Gb/s
(375MB/s) each with support for port multipliers (sitting on the storage
backplanes).
Ciao,
Dani
Michael Baldwin Bruce
is that right bruce?
Paul Ratcliffe
wrote:
> If you drive is SMART compliant (most are, these days) you can use
> "smartctl" to access the info:
>
> http://os2power.dnsalias.com/yuri/software/smartctl-b1.zip
Waste of time. It requires an outdated GCC runtime library from somewhere
and you have to have a 32 bit TCP/IP stack installed for some bizarre
reason. Binned it.
Stewart Buckingham
>>If you drive is SMART compliant (most are, these days) you can use
>>"smartctl" to access the info:
>>
>>http://os2power.dnsalias.com/yuri/software/smartctl-b1.zip
>
> Waste of time. It requires an outdated GCC runtime library from somewhere
> and you have to have a 32 bit TCP/IP stack installed for some bizarre
> reason. Binned it.
Can u suggest an alternative? I too still haven't managed to get
anything useful from this program, even with the old GCC runtime library
and the 32-bit TCP/IP stack :(
Stu/2
Paul Ratcliffe
wrote:
>>>http://os2power.dnsalias.com/yuri/software/smartctl-b1.zip
>>
>> Waste of time. It requires an outdated GCC runtime library from somewhere
>> and you have to have a 32 bit TCP/IP stack installed for some bizarre
>> reason. Binned it.
>
> Can u suggest an alternative? I too still haven't managed to get
> anything useful from this program, even with the old GCC runtime library
> and the 32-bit TCP/IP stack :(
http://hobbes.nmsu.edu/pub/os2/util/disk/smartmon.zip
is worth a look. I've put it into service here anyway - time will tell if
it lasts but it looks reasonably promising.
eric w
<ab...@orac12.clara34.co56.uk78> wrote:
>
> http://hobbes.nmsu.edu/pub/os2/util/disk/smartmon.zip
> is worth a look. I've put it into service here anyway - time will tell if
> it lasts but it looks reasonably promising.
this is SUPER!!!
eric w
"eric w" <er...@nospam.net> wrote in message
news:vSrfmdoFuNkL-p...@nospam.nospam.net...
Felix Miata
> There was a time when disk drive capacities were the same between
> the IDE and SCSI. Lately, however, IDE drives have been 80 GB,
> 120 GB, 160 GB, and 200 GB. I wonder why SCSI drives don't have
> the same capacities but just different interfaces?
Different size platters (smaller working diameter) for safe higher RPM
use and shorter stroke.
--
"Blessed is the nation whose God is the Lord." Psalm 33:12 NIV
Team OS/2 ** Reg. Linux User #211409
Felix Miata *** http://members.ij.net/mrmazda/
-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 100,000 Newsgroups - 19 Different Servers! =-----
glcm
wrote:
> >> 46877 hours = ~5.35 years...
> >
> > Definitely scheduled for "retirement" real soon now ;)
>
> Don't be so sure. Over the summer, I dragged out my first computers. It's
> an Epson QX-16 with a 20MB Seagate MFM drive. It booted up flawlessly, even
> though it's around 20 years old. Seagate knows how to make drives.
I wouldn't compare today's Seagate drives with their older ones.
Back in the day ALL drives were expensive beasts and Seagate was one
of the best, no doubt.
Heck, I think I still have a few ST-241's around that are running.
But now in days everything is about making it the cheapest and moving
product.
More now than ever, YMMV with regard to drives ...
Gregory L. Marx
tho...@antispam.ham
>>> Mike Ruskai wrote:
>>>> SCSI supports more devices. IDE has a max of two per channel. You can
>>>> string 15 IDE drives on a 16-bit SCSI chain with 15 of those devices.
>>> Which happens to be the very reason why both of them are dead-enders.
>> Please explain. Do you have a better alternative?
> Fast parallel interfaces (P-ATA, P-SCSI) are pleagued by the same
> serious deficits:
>
> 1) high number of connector pins
> a) reduces reliability
> b) increases variance of electrical conditions
How many connector pins does the Pentium 4 have?
In my experience, SCSI drives aren't any less reliable because of the
number of connector pins. Reliability would depend on the design of
the connector.
> 2) the "speed wall":
> all of the signalling lines need to reach stable values within a common
> sampling time "window". Faster transfer speeds narrow the window, while
> point 1b) above increases the probability of missing the window - even
> more so with longer cables and more units attached to the cable. Dealing
> with this problem increases system costs exponentially.
SCSI has Ultra 320 out. SATA is 150. No evidence there that a parallel
interface creates more difficulty in reaching high speeds.
> 3) compatibility with 1st generation hardware:
> the protocols have parts which need to be understood by all devices and
> need thusly to be carried out in the common denominator transfer mode -
> usually the slowest one. In the fastest implementations, the "1st
> generation mode" parts (arbitration, command delivery) eat up up to the
> half of the bandwidth available (worst case).
The Adaptec 29160 manual is quite clear about mixing older hardware
with newer hardware. You want the full speed, don't mix the slower
devices with the faster ones. Fortunately, the 29160 has four
connectors, one external LVD connector, one internal LVD connector,
one internal legacy wide SE connector, and one internal narrow
connector. So, it's not that hard to segregate your slower and
faster devices. And if you still have problems, you can always
install a second adapter and put your slower devices on a second
card. The backward compatibility could be considered a bonus.
> 4) buses: attaching multiple units
> a) increases problems 1b) and thus 2)
> b) potentially reduces the amount of bandwidth available to individual units
> c) requires arbitration of the shared medium,
> d) causes collitions, thereby increasing latency
The 29160 allows a total cable length of something like 12 meters
with lots of devices on it. I'm nowhere close to that length.
Doesn't look like a serious problem at this stage.
> The solution is serial interfaces like S-ATA and SAS:
>
> 1) as few pins as possible
> 2) the "speed wall" is much less paramount
> 3) drop support for legacy devices (or use bridges)
> 4) allow point-to-point connections only, with additional switching
> devices if required (f.e. SATA/SAS port multipliers).
So why do SATA drives top out at 150 right now?
> Have a look at the imminent future: PCI-X/PCI-e SATA/SAS (even
> bilingual) host adapters with 4/8 (or more) ports running at 3 Gb/s
> (375MB/s) each with support for port multipliers (sitting on the storage
> backplanes).
Well, I don't know what is in SCSI's imminent future, so I'll stick
with comparing what is currently available, and what I see is SCSI
at 320 and SATA at 150.
Daniela Engert
> Daniela Engert writes:
> So why do SATA drives top out at 150 right now?
Because 150 MB/s is plenty of headroom to fastest disk drives out. Being
faster than fast enough isn't better.
>>Have a look at the imminent future: PCI-X/PCI-e SATA/SAS (even
>>bilingual) host adapters with 4/8 (or more) ports running at 3 Gb/s
>>(375MB/s) each with support for port multipliers (sitting on the storage
>>backplanes).
>
>
> Well, I don't know what is in SCSI's imminent future, so I'll stick
> with comparing what is currently available, and what I see is SCSI
> at 320 and SATA at 150.
There is absolutely no point in discussing this. I've summarized the
reasons why the industry has chosen to get rid of parallel interfaces.
Your personal taste is - well - your choice.
Ciao,
Dani
tho...@antispam.ham
>>>>> Which happens to be the very reason why both of them are dead-enders.
>> So why do SATA drives top out at 150 right now?
> Because 150 MB/s is plenty of headroom to fastest disk drives out. Being
> faster than fast enough isn't better.
Sure, a single drive cannot saturate the interface, but multiple drives
doing simultaneous transfers can. Maxtor has some drives out capable
of sustained transfer rates of 98 MB/sec.
>>> Have a look at the imminent future: PCI-X/PCI-e SATA/SAS (even
>>> bilingual) host adapters with 4/8 (or more) ports running at 3 Gb/s
>>> (375MB/s) each with support for port multipliers (sitting on the storage
>>> backplanes).
>> Well, I don't know what is in SCSI's imminent future, so I'll stick
>> with comparing what is currently available, and what I see is SCSI
>> at 320 and SATA at 150.
> There is absolutely no point in discussing this. I've summarized the
> reasons why the industry has chosen to get rid of parallel interfaces.
> Your personal taste is - well - your choice.
Personal taste has absolutely nothing to do with it. The fact of the
matter is that if you're looking for maximum performance from multiple
disk drives, right now the choice is SCSI Ultra 320.
Michael Baldwin Bruce
Tonight's the night I shall be talking about of flu the subject of word
association football. This is a technique out a living much used in the
practice makes perfect of psychoanalysister and brother and one that has
occupied piper the majority rule of my attention squad by the right number
one two three four the last five years to the memory. It is quite remarkable
baker charlie how much the miller's son this so-called while you were out
word association immigrants' problems influences the manner from heaven in
which we sleekit cowering timrous beasties all-American Speke, the famous
explorer. And the really well that is surprising partner in crime is that
a lot and his wife of the lions' feeding time we may be c d e effectively
quite unaware of the fact or fiction section of the Watford Public Library
that we are even doing it is a far, far better thing that I do now then,
now then, what's going onward christian Barnard the famous hearty part
of the lettuce now praise famous mental homes for loonies like me. So on
the button, my contention causing all the headaches, is that unless we take
into account of Monte Cristo in our thinking George the Fifth this phenomenon
the other hand we shall not be able satisFact or Fiction section of the Watford
Public Library againily to understand to attention when I'm talking to you
and stop laughing, about human nature, man's psychological make-up some
story the wife'll believe and hence the very meaning of life itselfish
bastard, I'll kick him in the Balls Pond Road, bruce.
Hendrik Schmieder
>
> The solution is serial interfaces like S-ATA and SAS:
>
> 1) as few pins as possible
> 2) the "speed wall" is much less paramount
> 3) drop support for legacy devices (or use bridges)
> 4) allow point-to-point connections only, with additional switching
> devices if required (f.e. SATA/SAS port multipliers).
>
> Have a look at the imminent future: PCI-X/PCI-e SATA/SAS (even
> bilingual) host adapters with 4/8 (or more) ports running at 3 Gb/s
> (375MB/s) each with support for port multipliers (sitting on the storage
> backplanes).
>
> Ciao,
> Dani
Hello Dani,
if you speak about SAS do you mean Serial Attached SCSI ?
I haven't seen any SAS hostadapters or disk,
not speaking about optical drives like CD/DVD-ROMs or CD/DVD-Burners.
And there's the question if there will be OS/2 drivers for SAS
Hostadapters .
Hendrik
tho...@antispam.ham
>> Daniela Engert writes:
>>> Because 150 MB/s is plenty of headroom to fastest disk drives out. Being
>>> faster than fast enough isn't better.
>> Sure, a single drive cannot saturate the interface, but multiple drives
>> doing simultaneous transfers can. Maxtor has some drives out capable
>> of sustained transfer rates of 98 MB/sec.
> In practice or just by the specs?
You'd have to find somebody who has some to find out what the maximum
sustained throughput is in practice. The specs can be seen at:
> Which ones?
The Atlas 15k II.
> Can you show the output of
> Sysbench or tell me what else you do to measure this speed?
I do not have any of those drives. Mine are only 10k and an older
generation, but I can copy a hundred MB file in 4 sec. That's about
25 MB/sec. I just copied a 700 MB .wav file on my home computer in
29 sec, which is 24 MB/sec sustained.
>>> There is absolutely no point in discussing this. I've summarized the
>>> reasons why the industry has chosen to get rid of parallel interfaces.
>>> Your personal taste is - well - your choice.
>> Personal taste has absolutely nothing to do with it. The fact of the
>> matter is that if you're looking for maximum performance from multiple
>> disk drives, right now the choice is SCSI Ultra 320.
> To exceed the 150MB limit you need at least four of the above drives and
> copy stuff between each pair simultaneously.
Actually, two drives can do it. Transfers can be external to a single
PC when sitting on a Gbit ethernet.
> How often do you do that in real work situations?
Not sure what your point is here. If you wish to claim that such high
speeds are not needed that often, then you're arguing that parallel
interfaces are NOT a dead end, because they're plenty fast enough for
today's typical uses, and then you'd be agreeing with me that parallel
interfaces are not a dead end.
As for my personal uses, I'm dealing with CCD images for which a
single file is 700 MB in size. I purposely set up a multi-drive
system to avoid the thrashing that goes on when a single drive has
to reposition the head to widely separated cylinders to perform a
copy operation. On my home computer, I'm often dealing with
comparably-sized audio files, which I load entirely into memory
for editing purposes.
As for the future, I don't know what to expect. Mechanical disk
drives may be replaced by something like a flash card, or a
futuristic "data crystal". But it does seem reminiscient of the
RISC vs. CISC debate of several years ago. There were people
claiming that CISC was a dead end, that the performance of Intel
chips was going to top out. Well, that didn't come to pass.
Intel's engineers apparently had a few tricks up their sleeves.
I'm not in the industry, so I don't know what tricks the engineers
of parallel interfaces might still have up their sleeves. So
until I actually see parallel interfaces dropped in favor of
serial interfaces, I wouldn't be so fast to pound some nails into
the coffin.
Hendrik Schmieder
Stewart Buckingham schrieb:
>
> Wolfi wrote:
>
> >> Eide also isn't very cheap if you have to buy a new drive every 2
> >> years because the drive goes kaput.
> >
> >
> > Just a short correction for possible future use:
> > "kaput" is written correctly like this: "kaputt" ;-)
>
> My Chambers Twentieth Century Dictionary allows both spellings. Perhaps
> I should move into the twentyfirst century :)
>
> Stu/2
Just for the record.
There's the german noun "Kaput" which is a soldier coat,
but not used in today's language
And there's adverb "kaputt" , which means seriosly damaged,
so that it is not working anymore.
Source :
"Duden : Das große Wörtebuch der deutschen Sprache in 6 Bänden" ; Band 4
(1978)
Hendrik
Stewart Buckingham
Most English dictionary random access devices (books) allow for both
spellings.
Stu/2
tho...@antispam.ham
>>> In practice or just by the specs?
>> You'd have to find somebody who has some to find out what the maximum
>> sustained throughput is in practice. The specs can be seen at:
>>> Which ones?
>> The Atlas 15k II.
> OK, with that rotational speed it may actually get near 100 MB/s at
> times.
The claim is for a sustained rate, not just "at times".
>>> How often do you do that in real work situations?
>> Not sure what your point is here. If you wish to claim that such high
>> speeds are not needed that often, then you're arguing that parallel
>> interfaces are NOT a dead end
> No, I was trying to find out how you are filling the limit of your
> Ultra320 controller while I find it hard to even fill the SATA limit.
I don't have an Ultra 320 controller. First of all, I'm not aware
of one with an OS/2 driver, and second of all, you need a 64-bit
bus to avoid having the PCI bus be the limiting factor on speed.
I do have an Ultra 160 controller, however.
>> As for my personal uses, I'm dealing with CCD images for which a
>> single file is 700 MB in size.
> MegaPrime imaging?
Bingo.
> I though those were pipeline reduced only?
Under normal circumstances, if you're willing to wait. My project
requires less then 24-hour turnaround, thus we need to do it ourselves.
> But yes, I
> love those wide-field images, too. Just got new data from CTIO Mosaic2.
> :-) But because when you are dealing with them the CPU(s) usually also
> get a lot of work to do you will never see situations where the disk to
> RAM transfer is the bottleneck. At least that's my experience...
Every little bit helps, whether it's a faster CPU or a faster disk
drive. The biggest bottleneck right now is the network between
islands.
Daniela Engert
> Daniela Engert schrieb:
>
>>The solution is serial interfaces like S-ATA and SAS:
>>
>>1) as few pins as possible
>>2) the "speed wall" is much less paramount
>>3) drop support for legacy devices (or use bridges)
>>4) allow point-to-point connections only, with additional switching
>>devices if required (f.e. SATA/SAS port multipliers).
>>
>>Have a look at the imminent future: PCI-X/PCI-e SATA/SAS (even
>>bilingual) host adapters with 4/8 (or more) ports running at 3 Gb/s
>>(375MB/s) each with support for port multipliers (sitting on the storage
>>backplanes).
>
Yes, I do.
> I haven't seen any SAS hostadapters or disk,
> not speaking about optical drives like CD/DVD-ROMs or CD/DVD-Burners.
AFAIK there are SAS host adapters from LSI and Adaptec, and SAS disks
from Maxtor, but I might err on the actual manufacturers. I don't know
if this stuff is in the retail channels yet.
> And there's the question if there will be OS/2 drivers for SAS
> Hostadapters .
Implementing support shouldn't be too difficult, probably based on other
open source efforts.
By the way, I'm deeply impressed by a true hardware RAID solution which
requires *no* specific driver support (i.e. should work with DaniS506
with zero or nearly-zero additions) - but it's PATA or SATA: have a look
at http://www.netcell.com
Ciao,
Dani
Trevor Hemsley
Engert <dani%ngr...@nospam.de> wrote:
> By the way, I'm deeply impressed by a true hardware RAID solution which
> requires *no* specific driver support (i.e. should work with DaniS506
> with zero or nearly-zero additions) - but it's PATA or SATA: have a look
> at http://www.netcell.com
I looked. It sounds interesting but one of the most important parts was
missing... like how much!?
--
Trevor Hemsley, Brighton, UK.
Trevor-Hemsley at dsl dot pipex dot com
William L. Hartzell
Trevor Hemsley wrote:
> On Sun, 3 Oct 2004 16:15:47 UTC in comp.os.os2.setup.storage, Daniela
> Engert <dani%ngr...@nospam.de> wrote:
>
>
>>By the way, I'm deeply impressed by a true hardware RAID solution which
>>requires *no* specific driver support (i.e. should work with DaniS506
>>with zero or nearly-zero additions) - but it's PATA or SATA: have a look
>>at http://www.netcell.com
>
>
> I looked. It sounds interesting but one of the most important parts was
> missing... like how much!?
>
I agree. I could not find the product listed on xfxforce web site
<http://www.xfxforce.com/products.php>. The announcement was on
September 9, this year. Maybe they don't have anything yet.
--
Bill
Thanks a Million!