Windows 7 problem - hardware - software ?

[fohat.digs]

I have a large and robust system with what was high-end electronics 2-3 years ago, so probably well off the cutting edge now. All my software is up to date and I clean out the rubbish regularly.

All has worked properly, but boot times have been getting much longer in recent weeks. I have a small C:\ Samsung 830 SSD that is pretty much devoted to Windows 7 OS and almost nothing else. D:\ is a large conventional WD that is perhaps a little over half full.

I the last few days I have had a few random crashes and blue screens.

My first thought is that the SSD might be failing.

Any other opinions?

[Computer-Lab in Basement]

SSD failure is the most likely candidate.

Any error codes thrown up during the BSODs?

Also, how old is that Windows installation? Could just be your install is getting old and you need to re-format.

[SpAmRaY]

Have you tried checking the life of the drive using samsungs toolkit?

[IvanIvanovich]

When SSD fail, my experience is they simply disappear with a 100% failure without any warning. Performance degradation over time is not something uncommon. Most people don't leave it enough idle time for it to perform trim and garbage collection routines. I would suggest you grab partedmagic or similar and use it to do an ata secure erase. That will bring it back as close as is possible to the performance it had when new. Also, if you have not updated firmware for the SSD it is a good time to do that as well before reinstalling Windows.

[digi]

Is your SSD a Crucial by chance? I fixed a drive that was I thought might be dead (different than your issue) by upgrading the firmware on it.

[Computer-Lab in Basement]

Is your SSD a Crucial by chance? I fixed a drive that was I thought might be dead (different than your issue) by upgrading the firmware on it.

OP says it's a Samsung 830

[digi]

Is your SSD a Crucial by chance? I fixed a drive that was I thought might be dead (different than your issue) by upgrading the firmware on it.

OP says it's a Samsung 830

Doh! Well, maybe Samsung has firmware for the SSD also, dunno.

[physicsmajor]

Go get Memtest86 and run it overnight to check the RAM. I've had RAM go bad on me before, and it's a real challenge to diagnose. It's a rare failure mode, but well worth ruling out.

Also certain other hardware can fail and result in bluescreens if its driver was deeply enough integrated. This isn't as common in Win7 as in XP and prior, but it still can happen. If you recently added a new piece of hardware, consider removing that and seeing if the system is stable for a day or two.

[physicsmajor]

Also: if you have a 2-3 year old system, have you cleaned out dust inside with a vacuum or compressed air (or both) within the last 3-4 months?

That stuff builds up fast. My system is a similar vintage and I must do this about every 3 months or my temps creep up. My previous system had a GPU running idle temps in the 95C range because I failed to clean out the dust, and I learned my lesson...

Temps go down, fan speeds go down, it's a win win. Well worth your time, and this can definitely cause instability.

[fohat.digs]

Good point, but no.

This is a giant tower with something like 6 fans and very good airflow. I open it to change/add/shuffle hardware at least a couple of times a year, and thoroughly clean it each time.

The front fans have nice grille filters that seem to work very well, and these are intake.

I feel pretty sure that it is either a tired Windows installation or an SSD nearing the end of its useful life.

But the "nuke and pave" process is excruciating and daunting, that is why I rarely do it more often than about once every 2-3 years.

[dorkvader]

SSD failure is the most likely candidate.

I disagree.

At this point, there isn't enough information to know what's causing it. Go through the BSOD logs (with nirsoft's bluescreenview or something) and post the STOP codes and affected drivers here. I can use that to start to figure out what's wrong.

All the other ideas in this thread (cleaning it, RAM test, etc) are good ideas, but I feel this will help figure out what your actual problem is faster. Then we can figure out potential fixes. I feel this sort of diagnosing is better than trying random solutions until something happens (shotgun approach).

[fohat.digs]

diagnosing is better than trying random solutions

So, naturally, it has worked flawlessly for a week or so.

Last Tuesday it seemed that Windows 7 got a larger-than-average inoculation, so maybe there was a gremlin in my Microsoft/Windows installation somewhere.

At any rate, I am in no hurry, but I also feel that refreshing the OS every 2-3 years is probably worth it.

[fohat.digs]

And on a possibly-related semi-tangent, testing "strange" keyboards often produces bizarre results.

My standard setup is an F-122 running though a Teensy connected to a non-powered (good-quality name-brand) USB hub. In general, this works properly and boots consistently.

When I test a USB keyboard or mouse, I hot-plug it in to a vacant USB port. Often it works immediately, but sometimes a re-boot is required.

PS/2 keyboards are far more problematic. I assume that it is a given that they have to be recognized in a re-boot. But often, even with my standard "go-to" keyboards (a Model M, and a Northgate Omnikey and a Dell AT101W, both modded and transplanted with better switches), as well as many of the strange and obsolete new/old keyboards that I have bought, I often get boot problems, crashes, BSOD/"restore to earlier configuration" and other time-consuming and meaningless nonsense. Sometimes multiple iterations are required to get back to ordinary life.

But here is the weird part: with the computer completely "shut down" but not actually "off" via the hard-wire switch in the back, *sometimes* the PS/2 keyboard will cause it to boot by plugging-in, un-plugging, or sometimes even just pressing random keys!

WTF is that all about? Non-techie that I am, it almost seems like BIOS-level stuff!

[physicsmajor]

But here is the weird part: with the computer completely "shut down" but not actually "off" via the hard-wire switch in the back, *sometimes* the PS/2 keyboard will cause it to boot by plugging-in, un-plugging, or sometimes even just pressing random keys!

WTF is that all about? Non-techie that I am, it almost seems like BIOS-level stuff!

Your intuition is probably right. The only way what you describe could cause the tower to soft-power on is if the BIOS was looking for such a signal. Does the Num Lock or equivalent go out and stay dark on your PS/2 keyboards when the tower is "off"?

You could also look through your BIOS settings looking for something obscure that would probably sound like "Wake on _________." Might be next to "Wake on LAN," which allows the ethernet connection to wake the tower even while "off," a very convenient option for sysadmins.

[Novus]

So I remember this being an issue for some users with some mobos.
One thing you want to do is enable hot plug/hot swap under your bios settings (under SATA) for the port your SSD is on.