copying certain files to usb hangs PC?

[shorterxp]

I have a folder with files,
They take forever to copy to a USB drive. Explorer eventually hangs. Copy doesn't cmplete unless you wait an hour (lol).
If compressed beforehand, copies at normal speed (really quick).
More of a curiosity than an urgent problem... Any ideas?

EDIT: IT turns out that what OP describes is default behavior with files that are small enough in size. Alas, a leading indicator used in benchmark software is the transfer of 4kb or thereabout, sized files. Its completely normal.

[dencorso]

What're the versions of your machine's NTFS.SYS and zipfldr.dll?

[shorterxp]

5.1.2600.5512 for NTFS

zipfldr.dll shouldn't matter, the problem is copying files in a regular folder.
The files were not compressed with zip to begin with fyi, I just compressed them to test, i.e. when zipped, same files transfer faster than if they aren't. Explorer is doing something funny...

[shorterxp]

Bump

[mooms]

Do you have an antivirus ? It may be related.

[shorterxp]

No anti-virus running.

It worked fine before I re-did an nlite of SP3 to integrate AHCI drivers. Then I booted up that copy and made it main boot.

[shorterxp]

Skip

[shorterxp]

Tried alternative 'optimized' chipset drivers on fresh install of nlited XP, same thing happens.
Explorer goes unresponsive if you click cancel; the only way to get PC to respond is to unplug USB stick.
There are no reports of this particular USB stick failing online, so I'm left to assume its something to do with OS configuration and nlite.
Its just strange individual files hang explorer like this. Heres a video https://youtu.be/-n7ZTQh4Dhs

[shorterxp]

Power isnt an issue; removed all other devices from usb ports and transfers are still slow. Tried alternative USB Stick too.

According to on post on MS forums, Background Intelligent Transfer Service being may be related.
I should try another nlite/XP image with BITS included see if that changes anything first.

[patrickjburt]

Is it just the one usb drive or have you tried several?

[shorterxp]

It happens on all USB drives.

An alt. nlite install with BITS intact concludes its not culpable, however I suspect another, also absent service
Performance Logs and Alerts service, may be.

I shall make image with this installed and see. Expect report within a day.

[shorterxp]

Swapped motherboard for another and tried again, everything else is the same.

The files copy and eventually complete copying but it takes very long time, just not as long as it did on previous mobo.

Bloody weird I tell ya

[shorterxp]

https://arstechnica.com/civis/viewtopic ... 8&t=270700

[mooms]

So you swapped for a motherboard with the same chipset ?

[shorterxp]

Indeed. Latest bios on both.

Initially it was Intel based nForce 630i MCP73, Foxconn Irvine then I moved to AMD based nForce 710a MCP78, Pegatron Nutmeg. I was so pissed off at the constant crashes of the nutmeg it got smashed and Reverted to intel.
The only reason I like these is the plentiful features in mini-ITX format and that theyre cheap to populate.

I use XP due to the nature of the software I use.

I'm asking fernando on his forum to see if he can help. Within device manager for the AMD there was "nforce sATA controller" and driver was correctly showing. However on this intel, there is a Dual Secondary PCI IDE Controller with the Hardware ID of nforce SATA controller. No matter what .inf I apply to it, after restart theres still a yellow . If I use the setup.exe it it detects the sata cotnroller and the installer for that particular driver shows, but after restart its still the default XP one. Both boards have SATA controller options in bios. They were/are enabled.

I'm about to test a third, older board AMD sockethttps://en.wikipedia.org/wiki/NForce4#n ... F410.2F430 , Asus M2N6-AR.

[shorterxp]

Okay so loaded up new motherboard. Its more stable, files can be copied but verrrry slowly. If copy is cancelled it doesn't crash explorer unlike before. I'm using some pretty early nforce drivers, not fernandos ones.

I opened System Event Viewer and Cleared all events then copied files to USB to see what errors are logged afterwards:

Code: Select all

A parity error was detected on \Device\Ide\IdePort2.
The driver detected a controller error on \Device\Harddisk0\D.

[mooms]

I had a ASUS A8N Sli Premium (nForce 4 on S939) with Athlon 64x2, my first 64 bit and dual core CPU.
It was good on XP era, but not so much on Windows 7...

Before that, I had an Asus A7N8X-E Deluxe (nForce2 Ultra 400), and even before a MSI K7N 420 Pro (nForce), both on Socket A (462) with various Athlon XP.

I was a big fan of these NVIDIA chipsets, they were stable and performed well.

Then I switched on Intel, i5 2500K on ASUS P8P67, and now i7 6700K on ASUS Z170 Pro Gaming.

[shorterxp]

I didn't realise a USB stick consumed 300mA of power. There is a feature in Device Manager that lists each device consumption, to be sure I unplugged all but mouse and USB stick, the problem still persists.
I need to try an ALT power supply really. Current one installed is 250Watts, the default supply is 170 watts, I doubt it is the problem with this fact in mind.

[spawn]

as i recall the power issue can be exploited by wifi intrusion more specifically the motherboards power circuitry can be exploited to make the system switch on and off repeatedly. This happened to me on several older P4 and P4 multi and AMD motherboards. (buggy chips) where nvidia network ports reside one must take a time machine back to vista and windows 7 days for nvidia network driver that includes a layer to handle its in chip firewall that windows generic drivers do not do.
Parity errors can be caused by drivers. The newest driver isn't always the best. Most drivers are best on the motherboards own bundled set. Vista and windows 7 drivers will work on windows 8 and 10 just as well and if hardware was designed for vista or etc the same driver probably hasn't changed but bundled in a driver package with newer products with it.
I used windows XP X64 with my deluxe it was sluggish on 32bit. An external usb hub with mains power fixes power issues

[shorterxp]

IT turns out that what OP describes is default behavior with files that are small enough in size. Alas, a leading indicator used in benchmark software is the transfer of 4kb or thereabout, sized files.