Re: External hard drive problem

Tech-Archive recommends: Repair Windows Errors & Optimize Windows Performance

From: Pete (lansma_at_bigpond.com)
Date: 12/03/04


Date: Fri, 03 Dec 2004 14:32:44 GMT

Just for completeness-sake...strangely it seems this is a problem with
virtual memory. I changed the min and max values to 1.5 times my RAM
(512 Mb) and now the drive appears to work properly. I just did a
successful defrag, something I wasn't able to do previously using
Firewire. I will do some more testing tomorrow, but it looks promising.
The help came from the Dell forum. Yippy !

-- 
All the best,
Pete
------------------------------------------------
Home Page: http://users.bigpond.com/lansma
Location: 42°53'S; 147°19'E
R. McCarty wrote:
> I would visit the Drive web-site and pull the product specifications
> sheet (probably a .pdf) and check the "Rated" speeds. But yes the
> values you mention track with what I see on my USB 2.0 External
> Maxtor.  On my 5000-LE the access time is around 20 mSeconds
> and average Read/Write data rates are around 17-20 Megabytes.
>
> "Pete" <lansma@bigpond.com> wrote in message
> news:oRZrd.57309$K7.541@news-server.bigpond.net.au...
>> Thanks, you've been a big help. I have tried to alter the policies
>> without a change in the problem. I have just posted to the Dell
>> peripherals forum. I don't think I am alone in having firewire
>> problems it seems.
>>
>> One FINAL question: do the read/write speeds I reported sound within
>> normal limits?
>>
>> Again, many thanks.
>>
>>
>> --
>> All the best,
>> Pete
>>
>> ------------------------------------------------
>> Home Page: http://users.bigpond.com/lansma
>> Location: 42°53'S; 147°19'E
>>
>>
>>
>> R. McCarty wrote:
>>> I wasn't aware that your PC was a notebook. In that case, then
>>> both USB ports are 2.0 (High Speed).
>>>
>>> As to the Firewire connection experiencing Delayed write failures
>>> it could be due to the Polices on the drive. I have a Maxtor
>>> external USB drive. Whenever it is connected, it's Device Manager
>>> policy is set to "Optimize for Quick Removal" which doesn't use
>>> delayed writes. You should connect via both technologies and go to
>>> Device Mgr & click the Policies Tab and see how the caching is set
>>> in each setup. Perhaps when you use Firewire, the disk cache can't
>>> flush data fast enough and that's why you are getting the error.
>>>
>>> "Pete" <lansma@bigpond.com> wrote in message
>>> news:YEXrd.57179$K7.9164@news-server.bigpond.net.au...
>>>> Thanks for that, but I remain confused. I have a Dell notebook and
>>>> only have two USB ports. Doers this mean they are both USB 2.0
>>>> based on the info I provided above or just one of them?
>>>>
>>>> I have done some tests of read/write/and random access speeds for
>>>> the USB 2.0 (?) port versus firewire. (It is only when I use
>>>> firewire that I get the "Windows - Delayed Write Failed" problem).
>>>> I'd welcome your feedback:
>>>>
>>>> Firewire
>>>> ----------
>>>> Sequential Write: 22.4 Mbps
>>>> Sequential Read: 29.4 Mbps
>>>> Random Access: 16.7 Mbps
>>>>
>>>> USB
>>>> ----------
>>>> Sequential Write: 23.3 Mbps
>>>> Sequential Read: 24.3 Mbps
>>>> Random Access: 12.8 Mbps
>>>>
>>>> Are these fairly typical rates?
>>>>
>>>> Thanks a million.
>>>>
>>>> --
>>>> All the best,
>>>> Pete
>>>>
>>>> ------------------------------------------------
>>>> Home Page: http://users.bigpond.com/lansma
>>>> Location: 42°53'S; 147°19'E
>>>>
>>>>
>>>>
>>>> R. McCarty wrote:
>>>>> Yep - That's how most Intel Chipset 82801 Controller Hubs are
>>>>> enumerated in Device Manager.  Your PC should have 4 fixed USB
>>>>> ports on the back plane, then two separate connectors. One for use
>>>>> with front panel USB ports and the other to a PCI slot access.
>>>>> Unfortunately, there isn't a software tool for diagnosing which
>>>>> type of USB (Full -1.1 or HighSpeed 2.0) are present. Adaptec has
>>>>> one called USB Control but it only works with Adaptec PCI Add-in
>>>>> USB cards.
>>>>>
>>>>>  Also, the ports aren't logically arranged - You need to change
>>>>> the Device Manager default view to "View by Connection", then
>>>>> expand the ACPI, ACPI Compliant System, PCI Bus, USB hubs and you
>>>>> can
>>>>> see how actual devices are connected to which HUB by their
>>>>> identifier (24C4, 24C2, etc). In other words USB ports 1 & 2, 3 &
>>>>> 4, 5 & 6 don't always equate to the same Hub. It's confusing, but
>>>>> helps if you have a heavily loaded HUB where power limits or
>>>>> bandwidth are
>>>>> an issue.
>>>>>
>>>>> "Pete" <lansma@bigpond.com> wrote in message
>>>>> news:U2Ord.56559$K7.27512@news-server.bigpond.net.au...
>>>>>> Thanks R.,
>>>>>>
>>>>>> In Device Manager under Universal Serial Bus controllers I have 8
>>>>>> entries:
>>>>>>
>>>>>> Intel(r) 82801DB/DBM USB 2.0 Enhanced Host Controller - 24CD
>>>>>> Intel(r) 82801DB/DBM USB Universal Host Controller - 24C2
>>>>>> Intel(r) 82801DB/DBM USB Universal Host Controller - 24C4
>>>>>> Intel(r) 82801DB/DBM USBUniversal Host Controller - 24C7
>>>>>> USB Root Hub
>>>>>> USB Root Hub
>>>>>> USB Root Hub
>>>>>> USB Root Hub
>>>>>>
>>>>>> and if I plug my external drive in I get another entry:
>>>>>>
>>>>>> USB Mass Storage Device
>>>>>>
>>>>>> Does this smean my 2 x USB ports are 2.0?
>>>>>>
>>>>>> Thanks.
>>>>>>
>>>>>>
>>>>>> --
>>>>>> All the best,
>>>>>> Pete
>>>>>>
>>>>>> ------------------------------------------------
>>>>>> Home Page: http://users.bigpond.com/lansma
>>>>>> Location: 42°53'S; 147°19'E
>>>>>>
>>>>>>
>>>>>>
>>>>>> R. McCarty wrote:
>>>>>>> If the device is USB 2.0 capable and you plug it into a USB full
>>>>>>> speed port (Version 1.1) XP will report that a USB2.0 (High
>>>>>>> Speed) device is being used in a non-USB2 port.
>>>>>>>  Determine USB capability by the descriptor in Device Manager
>>>>>>> for Universal Serial Bus Controllers by checking for "Enhanced
>>>>>>> Controller" - which denotes USB Full Speed 2.0.
>>>>>>>  For benchmarking, I would use DiskSpeed 32 which can display
>>>>>>> the important aspects of Burst, Sustained transfers and access
>>>>>>> time - which shouldn't vary between Firewire and USB.
>>>>>>> Disk Speed can be downloaded from:
>>>>>>> http://www.geocities.com/vgrinenko/DiskSpeed32/
>>>>>>>
>>>>>>> "Pete" <lansma@bigpond.com> wrote in message
>>>>>>> news:2oMrd.56478$K7.18397@news-server.bigpond.net.au...
>>>>>>>> Now, here's the thing:
>>>>>>>>
>>>>>>>> When connected using Firewire, I get the problems reported when
>>>>>>>> trying the defrag but using USB 1.1 I don't. Well, at least I
>>>>>>>> think it's 1.1 but it doesn't seem much slower than Firewire,
>>>>>>>> if at all, which I think is strange.
>>>>>>>>
>>>>>>>> Does this offer further clues?
>>>>>>>>
>>>>>>>> Also, is there a way to check whether I have USB 1.1 vs USB
>>>>>>>> 2.0?
>>>>>>>>
>>>>>>>> Finally, is there a way I can properly test the speed of the
>>>>>>>> drive running USB versus Firewire - like a diagnostic tool I
>>>>>>>> can download somewhere?
>>>>>>>>
>>>>>>>> Thanks again.
>>>>>>>>
>>>>>>>> --
>>>>>>>> All the best,
>>>>>>>> Pete
>>>>>>>>
>>>>>>>> ------------------------------------------------
>>>>>>>> Home Page: http://users.bigpond.com/lansma
>>>>>>>> Location: 42°53'S; 147°19'E
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Alex Nichol wrote:
>>>>>>>>> Pete wrote:
>>>>>>>>>
>>>>>>>>>> "Windows - Delayed Write Failed
>>>>>>>>>> Windows was unable to save all the data for the file G:????.
>>>>>>>>>> The data has been lost. this error may be caused by a failure
>>>>>>>>>> of your computer hardware or network connection. Please try
>>>>>>>>>> to save this file elsewhere."
>>>>>>>>>>
>>>>>>>>>> I'm reasonably clueless about what to do in these situations
>>>>>>>>>> but I did run the Windows XP disk error checking utility
>>>>>>>>>> (which took literally a couple of hours). This did not report
>>>>>>>>>> any problems.
>>>>>>>>>
>>>>>>>>> The best thing in such a case is to turn off the delayed
>>>>>>>>> write - the loss is not in practice very much and better have
>>>>>>>>> it work reliably. Go to
>>>>>>>>> Control Panel - System - Hardware - Device Manager
>>>>>>>>> look for the drive, double click and in the Policies page
>>>>>>>>> uncheck 'Enable Write caching'
>>>>>>>>>
>>>>>>>>>  I would also go to the USB controllers section,  lower down
>>>>>>>>> and look for the  USB Root Hub that shows as being in use for
>>>>>>>>> this device on its Power page.  Then in its Power Management,
>>>>>>>>> uncheck 'Allow computer to turn off to save power', which may
>>>>>>>>> well be the underlying cause


Relevant Pages

  • Re: External hard drive problem
    ... > values you mention track with what I see on my USB 2.0 External ... >> R. McCarty wrote: ... Whenever it is connected, it's Device Manager ... Sequential Write: 22.4 Mbps ...
    (microsoft.public.windowsxp.general)
  • Re: HD and Digital Cable for America 2004. or at least MCE 2005!
    ... 480 mbps for USB 2.0 vs 400 mbps. ... but USB 2.0 is everywhere now. ... The latest Firewire is 800 mbps. ... the firewire port and was wondering what it was for. ...
    (microsoft.public.windows.mediacenter)
  • Re: Did Apple design Firewire?
    ... The speed statistic published for USB 2 is a BURST speed. ... isn't actually capable of maintaining 480 mbps. ... The speed of Firewire 400 is a sustained rate. ... Licensing fees are what is killing Firewire. ...
    (comp.sys.mac.advocacy)
  • Re: Epson 4990 firewire problem
    ... with the USB 2.0 High Speed interface it increases a little bit. ... the theoretical speed of a high speed USB 2.0 interface (480 Mbps ... software reports the scanner isn't responding. ...
    (comp.periphs.scanners)
  • Re: External hard drive problem
    ... then> both USB ports are 2.0. ... it's Device Manager policy is> set to "Optimize for Quick Removal" which doesn't use delayed writes. ... >> Sequential Write: 22.4 Mbps ... USB Root Hub ...
    (microsoft.public.windowsxp.general)