Jump to content
Covecube Inc.
  • 0
Fred Leggett

Scanner unfreezing my system.

Question

This doesn't really have anything to do with Scanner, but it IS related, so please keep reading.

Every once in awhile, my Windows 10 system will partially freeze.  By partially, I mean I can still access the start menu and run certain programs, but the rest of the system is unresponsive.  Bizarrely, this seems to happen only when I'm watching YouTube videos.  I think it might've also happened once when I was running a video through MPC-HC.  I also witness the occasional BSOD caused by a core video driver, which might be related and is distressing, but can't be heat-related (see below).

For a few weeks now, Scanner has been reporting that my NVMe (a Samsung 950 Pro) is running seriously hot.  I finally had enough of that and bought an active SSD cooler.  Once installed, temps dropped considerably.  Scanner still reports a warning, but only a degree above spec instead of the almost 10 degrees at which it was running before.

Unfortunately, that hasn't stopped the partial freeze issue.  I've checked my CPU temps and they're all nominal.  In fact, they're all exceedingly low, as I only use this box for HTPC duties and use a Corsair AIO.

This is where Scanner starts to really enter the picture.  When the partial freeze occurs, I can invoke Scanner and it will unfreeze the freeze.  I originally thought Scanner might be the problem, but the freezes persisted even after uninstall.  So, as it turns out, Scanner is actually the one program I can rely on to restore my system to a usable state.

My question is this - what is Scanner doing to unstick my system?  Is it sending some sort of wake or initialization command to my drives when invoked?  If I can understand this better, I might have a shot at determining the real culprit.  While I'm not running the most complex of setups, I would rather avoid having to reinstall Windows and everything else if I can help it.  Plus, there's no guarantee that will fix the problem.

Thanks in advance.

Share this post


Link to post
Share on other sites

11 answers to this question

Recommended Posts

  • 0

Well, it's querying drives, pretty regularly. But it does that automatically, so you wouldn't need to open the UI for this.

However, the UI does connect to the "performance counters" on the system, and that queries a bunch of parts on the system.

My guess though, would be a GPU related issue, though.  A bad driver, or maybe a damaged GPU. 

Share this post


Link to post
Share on other sites
  • 0

I'm beginning to suspect the CPU.  I'm using the onboard graphics of my i5-4960K, which contains Intel HD graphics 4600.

I don't really know why the CPU would be the culprit, though, as it seems to work fine in all other respects.  And, as I wrote earlier, the temps are all fine, so it can't be an overheating issue.  Drivers are all up-to-date.  Windows is fully updated.  The BIOS is the latest.

Very strange.  I guess I'm gonna have to do additional troubleshooting.

Share this post


Link to post
Share on other sites
  • 0

My apologies for not getting back to this.  No good reason why, just the days are whizzing by.

I actually do have the Intel RST drivers installed.  The latest ones I could find at the time (15.9.4.1041).  I'm also using a pair of Supermicro/Marvell SAS controllers that have long been superseded by more advanced controllers.  I could just use the packaged Windows 10 drivers, but I absolutely have to use the Marvell ones.  Have you heard of any problems with Supermicro cards?  The product page is https://www.supermicro.com/products/accessories/addon/AOC-SASLP-MV8.cfm

Share this post


Link to post
Share on other sites
  • 0

Well, for the intel RST, you could probably safely remove the drives.

As for that Marvell based card, yeah, I've seen a couple issues that were ultimately caused by this card.  Replacing it with an LSI based card fixed the issue. 

In fact, I'm not a fan of marvell based cards, in general. 

Share this post


Link to post
Share on other sites
  • 0

I've had those Marvell cards for awhile now.  At the time, they were pretty popular, as they were a VERY cheap way to obtain SAS connectivity and increase raw port count without needing a multiplier.  I'm using two inside a Norco 4224 case alongside my motherboard's SATA headers.  They run EXTREMELY hot, so I'm thinking of replacing the heatsinks with an active cooling solution.

My dream was to one day obtain one of these monsters:

https://www.newegg.com/Product/Product.aspx?Item=N82E16816151173

but I never was in a financial position to buy one.

Well, I guess I'll remove the RST drivers to see if that solves the problem.  The thing is I can't trip these semi-freezes manually, so I might never know if that was the issue.

Share this post


Link to post
Share on other sites
  • 0

Yeah, but with hardware, you absolutely get what you pay for...

That said, I know we've had some issues with some areca cards, getting SMART data and the like, IIRC. So ... I'd really recommend an LSI/Avago/whatever-it-is-now card. Those seem to be the most stable SAS HBA cards.

Especially when something like this is an option:
https://www.newegg.com/Product/Product.aspx?Item=N82E16816118282

:)

 

And for the system freeze:
http://wiki.covecube.com/StableBit_DrivePool_System_Freeze

It may be a good idea to do that, to get a memory dump of the system. 

Share this post


Link to post
Share on other sites
  • 0

Hmmm.  Someone wrote a pretty meh review of that Avago card you linked.  I really wish these cards would drop in price some. I would happily pay around $500 for a brand-name model with six 8643 connectors, but $1k is just out of my price range.  I don't like going the used route, but I'm beginning to think it might be the only way to replace my current Marvell pair with something less sketchy.

That system freeze memory dump procedure looks like it would generate some extremely interesting results.  Thanks for the link.  If the Intel RST drivers are the culprit, that would be invaluable to know.  However, if it turns out to be Marvell driver related, I'm really screwed.

Share this post


Link to post
Share on other sites
  • 0

That review was based solely on ATTO (synthetic) benchmarking, and should be ignored. 

Though, I don't have hands on for one of these cards, so it's hard to tell. 

But if you are willing to go SAS2 (which is MORE than fast enough), you can get a LSI SAS 9207-8i, and a SAS Expander for $200-300, give or take.  And that would get you 24 ports (or 40, if you get 2 expanders).   It's basically what I use, actually. 

 

And yeah, the memory dump is a great way for diagnosing those sort of issues, if you know how to read them. 

Share this post


Link to post
Share on other sites
  • 0

I'll investigate the port multiplier route, but I wanted to limit the amount of raw cards in my system, hence my attraction to the Areca and similar options.

I'm conducting an experiment.  Instead of invoking the Scanner interface when the semi-freeze occurs, I'm letting it run all the time.  It's been running like that for the past two days and I haven't experienced any freezes.  If I make it to the 10-day mark, I'm going to assume Scanner is somehow defeating or bypassing the problem.  If it gets there, do you have any ideas what could be happening?

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×