Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

I have this peculiar problem that is like "disk thrashing on steroids", that is not technically thrashing since it's not the lack of RAM that is causing it and it can even happen on times of very low resources usage. However, it can hang the system for a few moments, up to 10 seconds sometimes if the application reads from it. It also does not appear to hang the OS itself, but the application that reads from it

That is:

  • It happens on videos that are read from the internal HDD and not if they are read from an external USB drive. The video hangs as well as the audio and it resumes later. The rest of the OS does not appear hang completely, but applications such as browsers also "pause" momentarily.

  • It happens on games provided they are not installed on an external USB drive.

  • It happens on browsers and other applications that are installed on the internal HDD.

The computer in an HP DV7-5000 with the product number XE385EA. "Relative" models had an SATA problem and were eligible for replacement but, I heard that did not apply for the "HM55" chipset and it appears this model has the HM55 chipset so it's not that problem, even if I'm not completely sure they didn't screw up with that information.

I search for 'thrashing' and HM55 and I don't see anything definitive. Or with "pause" "hm55".

The problem appears also on completely different OSes (Linux, Win7 and Win8) so it appears closer to hardware, though you never know if both have the same software problem (unlikely).

Any idea what to do?

I wonder if a new internal disk would fix it or it's a controller problem.

The problem is that the laptop does not have a 2nd slot for a disk (not even connections) so one can't just add a second.

It occurs once every few minutes (7-10 minutes) for a few seconds and it was occurring since the beginning.

The disk led is on when it happens.

I have some limited programming knowledge: it reminds of an fread() hanging a single-loop app, but in a deeper hardware sense since it occurs on apps that provide for that.

The disk/controller from CrystalDiskInfo:

-- Controller Map ----------------------------------------------------------

  • Intel(R) Mobile Express Chipset SATA AHCI Controller [ATA]

    • ST9500420AS

    • hp CDDVDW TS-L633R

Hrm, I wonder if disabling the cd will fix it. I doubt it, but I may try it.


Additional thoughts I wrote in an HP forum:

With both caching options selected the problem is alleviated (both boxes ticked in the relevant device manager properties page for the drive). It still occurs but it's more rare and it appears to last for fewer seconds.

With caching options disabled (both caching options unticked in device manager for the drive) it occurs much more frequently.

It doesn't appear to be much different between normal and safe mode, though it may be related to frequency of operations/activity in general of writes or reads.

I've been trying to figure out what it ("thrashing+pausing apps installed on the same drive") means and it's tricky, though I suspect it's related to the 69 reallocated sectors reported by S.M.A.R.T. They are stable for years so there doesn't seem to be deterioration (and Seatools report health OK) though I suspect they initially occured in critical sections that force a reparking.

However, it may be completely unrelated to that since the pattern is relatively irregular since the only predictable pattern is that the higher the activity of the drive the more frequently in occurs. Hrm, unless it's a sort of logging activity in S.M.A.R.T. logging areas, but those are probably of a solid state nature, I assume.

One of the trickiest parts about the situation is that it's not clear that it's drive and not controller related, so another drive may have the same issue.

And it appears to be so closely related to hardware that I've seen it once "thrashing" in the same way while booting (when still on BIOS info stage) with no OS at all. It did delay the system even then IIRC.

What is certain is that it doesn't penetrate the CPU and overal system since running anything from an external USB drive from the OS it never pauses it (if it's installed/run on the external drive). It's strictly something between the drive and up to the controller or bus. No further. And it's likely it's not software related, at least not initially (since it can be alleviated with less software stress but not completely).

share|improve this question
add comment

2 Answers 2

It may be the drive has a bad area and periodically is retrying when it gets to that spot. Look for a disk utility to test the integrity of you drive ASAP. You can try and look to the manufacturer of the hard drive for a free version. Click Start, right click Computer, go to Properties, go to Device Manager and look for Disk drives. Identify your drive (Western Digital or Seagate) and then go to their website. They will usually have a utility for free to test your drive. Dependent on the drive, look for a non-destructive test. May take a LONG time to run. Back up your data.

share|improve this answer
    
The drive is like that for more than 2 years and I have made checks. The only suspicious thing is that (for years) the 'reallocated sectors count' is stuck at 96 which is theoretically suspicious but even if it's the problem it doesn't escalate since the number is stable since the beginning. –  leladax Jan 17 '13 at 14:50
add comment

[Self post] I have found a way to greatly alleviate the problem. I don't know if it's solved completely but it's certainly eased: Disable Windows write-caching from the drive's options in control panel (yes, tick both boxes there). Also according to Intel Rapid Storage's documentation their service should aid in performance when that option in used. In fact they appear to hint to users on the main window of their app to disable windows' caching even if they warn it increases the risk when unexpected power failures occur (this may hint to the whole thing being a deficiency of the controller).

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.