Windows 7 x64 Ultimate Edition Hangs every so often

[Acidflare]

Hi everyone, So my I am having a small issue with my windows 7 installation, I have a USB 3.0 UASP Dual-Bay 2.5"/3.5" HDD Dock that I use as an external storage device it is hooked up to the notebook with a startech.com usb3.0 with UASP 2 port flush mount 54mm Expresscard.
one of the HDD's is my library for steam and the other drive is my external storage drive some times when there is nothing going on but my AV running my one of my HDD's will start to read/write(There's indicators on the dock) and i'll notice that i'm getting a virus definition update, weird that uses my externals which in the device manager are labeled as USB devices the HDD model is listed with USB after it for example my WD 1TB WD10EZEX will be listed as "WD10EZEX USB" in device manager and in the hardware tab when you click properties on any drive.

now the funny thing is sometimes when the definition update is done updating it will hang and I have to unplug the dock to get it to work again, any attempt to use the HDD's results in windows explorer freezing up forcing me to unplug the dock anyways to get windows to operate properly again.

Earlier on this happened when I was downloading Ultra Street Fighter IV, I was watching a movie(to kill the wait time) off the 1TB and USF IV was being downloaded to the 250GB Steam Library drive my media player froze and windows became unstable until I unplugged the drive.

I've experience this issue with a USB 2.0 drive before as well but only when ejecting and trying to restart windows, never in the middle of doing operations.

I have the windows 7 xHCI drivers for the dock and the expresscard installed i'm pretty sure I have everything configured properly so I am completely unsure of why this occurring maybe someone else can shine some light on this pretty dim issue, it's kind of annoying to unplug 2 drives and risk data corruption on not 1 HDD but 2 HDD's

 

[Originality]

The first thing that comes to mind with this is an IO read error, probably from bad sectors. I suggest running CHKDSK to scan for bad sectors and maybe clean up any orphan files.

 

[Acidflare]

The first thing that comes to mind with this is an IO read error, probably from bad sectors. I suggest running CHKDSK to scan for bad sectors and maybe clean up any orphan files.

the xHCI driver for the dock uses an S.M.A.R.T program called iSMART that tells me if there are any disk error's the 1TB is fairly new only a few months old although the 1TB has more Sectors Reallocated then the 250GB which is older, the 250GB has 36 sectors Reallocated and the 1TB has 120 sectors reallocated. I know the 1TB has a lot of files that need to be organize, think this is what might be causing some issue?

 

[Originality]

The thing about SMART is that it uses a detection and reallocation algorithm that works gradually, in the background. If you're having problems now, that's when running a CHKDSK is useful - it forces a full scan of the drive that will detect and fix any bad clusters that the algorithm hasn't found yet.

The HDD isn't the only possibility. There could also be an issue with the dock's controller, or even your motherboard's USB controller/driver stack. It could be an anti-virus or windows indexing or dropbox attempting to scan the drive, building up long access queues, and choking up any system calls/interrupts until one of the microprocessors in the chain manages to work through its cache. Unfortunately I'm not an expert in these areas to be able to suggest how to definitively track down these possibilities outside of simply running CHKDSK (which tends to fix 80% of problems on healthy HDDs).

EDIT: when in doubt, request an insight from someone who knows more (e.g. FAST6191).

 

[Acidflare]

Yeah I agree FAST6191 knows his stuff, there is 2 different xHCI drivers installed, one set for the dock and the other set for the USB 3.0 expresscard, I think it's something to do with windows and having a drive sit on USB for so long, it doesn't happen all of the time, But it can be frustrating when your trying to do something and it happens.

I didn't know that about S.M.A.R.T. I guess one night i'll have to run chkdsk when I actually get tired at a decent time.

 

[Originality]

One more idea that just came to mind, maybe it's just a spindown issue. Drives that haven't been accessed for a while tend to spindown to save energy and wear. When they next get accessed, they have to spin up again before any command queues get processed. That can cause a freeze, but usually no longer than 15 seconds or so.

 

[Acidflare]

One more idea that just came to mind, maybe it's just a spindown issue. Drives that haven't been accessed for a while tend to spindown to save energy and wear. When they next get accessed, they have to spin up again before any command queues get processed. That can cause a freeze, but usually no longer than 15 seconds or so.

The dock has a spin-down timer coded into the board I think, if the drives haven't been accessed for x amount of time the drive access indicators will turn off and the drives will enter a low power spin mode. I haven't figured out what the time out is yet, too lazy really to activate the HDD's and use a stopwatch on my android to see how long it is. but that isn't the case though but I do understand your theory. my issue happens when the drives are being accessed, I believe it's the HDD Cache filling up and not dumping properly

 

[Acidflare]

So I think I fixed this issue.

I found out that without ASPM or PCI-e LPSM set to at least moderate that the CPU will always send dummy data through the PCI interface when no data is actually being processed to keep the connection stream alive for consistent access time. Now don't forget my HDD's are on a USB 3.0 with UASP 54mm ExpressCard which is in fact the mobile form of PCI-e x16 apart from mPCI-e which is also PCI-e x16.
my notebook actually has 3 PCI-e x16 slots, The 54mm express card slot, the mPCI-e slot that my WLAN card is installed to, and I have an actual graphics card custom made by toshiba that's on a PCI-e x16 slot.

Also my express card uses a renasas uPD72020 chipset so if I go to Start>Devices and Printers I see the express card there and can go to properties go to the hardware tab double-click on the hub device it opens a new window that has a Firmware Information Tab(The tab it opens up) and a power management tab that has a tick box for disabling USB 3.0 Power Management Functions.

So by setting the PCI-e Link Power State Mangement option to moderate power savings and turning on usb 3.0 power management functions(I had them turned off for performance) it seems to have fixed the issue but only time will tell if it actually works.

I woke up this morning and saw that the drives were reading but nothing was going on but my AV updating definitions again. So I unplugged the drives and felt them boy were they hot I possibly could have gotten a small burn if I hand held my finger on it for a good minute. I turned off the dock let them cool down and realized "Hey wait 2 months ago I turned off all power management functions for USB and PCI-e" i'm not having any other USB device issue's I have 2 usb flash drives connected for readyboost they don't fail on me like the HDD's do so I didn't touch the usb selective suspend it's still turned off hopefully this fixes my issue, I'd rather have less drive performance for the fact that it doesn't lock up and make me have to unplug 2 drives risking data corruption.

If only Linux could play my games that I play on windows I wouldn't have this issue there is no data corruption in linux, but sadly DX doesn't work to well with linux, i'm glad steam is putting out some linux games though