James wrote:
> both computers are pretty new i think they should be able to support
> transfering files just fine with their hardware. im guessing either a NIC
> settting is the case here or a bad NIC. this NIC im using is the onboard one
> that came with this msi 865PE neo-2 ls motherboard. any nic hardware
> settings u can think of that would cause this?
>
> "Bob Willard" <(E-Mail Removed)> wrote in message
> news:uO%(E-Mail Removed)...
>
>>James wrote:
>>
>>>what things could be differn't. both are clean installs of xp sp1
>>>
>>>
>>>"Bob Willard" <(E-Mail Removed)> wrote in message
>>>news:uEr$(E-Mail Removed)...
>>>
>>>
>>>>James wrote:
>>>>
>>>>
>>>>
>>>>>but lots of files tranfer fast when i use another workstation to the
>>>
>>>same
>>>
>>>
>>>>>server.
>>>>>
>>>>>
>>>>>
>>>>>"Bob Willard" <(E-Mail Removed)> wrote in message
>>>>>news:(E-Mail Removed)...
>>>>>
>>>>>
>>>>>
>>>>>>James wrote:
>>>>>>
>>>>>>
>>>>>>
>>>>>>>how come my transfers are so slow when copying files from my xp
>
> machine
>
>>>>>to
>>>>>
>>>>>
>>>>>
>>>>>>>my win2k machine on my 100mbps switched lan?
>>>>>>>large file = 100MB .mdb
>>>>>>>small files = total 7MB's of 400 text files
>>>>>>>
>>>>>>>it takes 12 mins to transfer the 7 MB's and only 30 seconds to
>
> transfer
>
>>>>>the
>>>>>
>>>>>
>>>>>
>>>>>>>large file.
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>And how come digging a ditch with one backhoe is faster than using
>>>>>>ten men with shovels? Or using a thousand men with teaspoons?
>>>>>>
>>>>>>Files are transferred one at a time. For each file, the source PC
>
> must
>
>>>>>>look up the file's metadata (to translate the file name to the file
>>>>>>location, check security params, etc.). Worse still, is what the
>
> target
>
>>>>>>PC must do for each file: it must find some unused space on the HD,
>
> then
>
>>>>>>allocate some of that space, then create the metadata for the new
>
> file,
>
>>>>>>then ask the source PC for the data, then copy the data to the
>
> allocated
>
>>>>>>space, then update the metadata (when all data has been copied), then
>>>>>>notify the source PC that the copy is done.
>>>>>>
>>>>>>It takes a lot of work to copy a file across a net, whether the file
>
> is
>
>>>>>>tiny or huge. For a huge file, copying data take most of the time;
>
> for
>
>>>a
>>>
>>>
>>>>>>tiny file, copying data takes far less time than all of the other
>>>
>>>overhead
>>>
>>>
>>>>>>work.
>>>>>>--
>>>>>>Cheers, Bob
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>From your question, I assumed that you were interested in why bunches
>>>>of files take longer than a single file to transfer; that is the issue
>>>>I addressed. If you want to understand why one PC is faster than
>
> another,
>
>>>>then you should conduct some measurements in which only one variable at
>>>>a time is changed -- and "the workstation" is not one variable, it is
>>>>probably many.
>>>>--
>>>>Cheers, Bob
>>>>
>>>
>>>
>>>
>>NIC speed, RWIN, protocols used, DNS settings, HD speed, CPU frequency,
>>IDE speed, RAM speed and quantity, pagefile min/max params and HD(s),
>>XP PRO v. HE, NTFS v. FAT32, user credentials, other traffic on the LAN,
>>HD RPM, HD fragmentation, CPU type, cache params, MB type/revision, BIOS
>>type, version, and settings, NIC vendor/model, NIC card v. chip,
>
> concurrent
>
>>apps running, domain v. workgroup, etc.
>>
>>And, maybe, lots of other stuff. It takes care and effort to compare two
>>things that are *exactly* alike except for one variable.
>>--
>>Cheers, Bob
>>
>
>
>
10 Mb/s, either explicitly set or the result of a bad auto-detect.
Or, HDX instead of FDX, explicit or bad auto-detect.
And, I would look at the DNS settings and protocols used for diff's
between those two PCs. Also, make sure you have the same credentials
and the same filesystem (preferably NTFS) with the same file sharing
(simple or ACLs) on both.
--
Cheers, Bob