HDD cache problem?

[rattyocaster]

Recently I spoke with someone who said his computer kept freezing without any real warning and after doing a proper reboot the system would seem fine. I checked the error logs and came up with nothing and I ran hard disk and memory checks and they all checked out ok. However, at the time I was having similar issues with a PC of mine which had an SSD in it, then the Windows 10 upgrade came out and that pretty much killed my SSD to the point where the system will freeze every 5 minutes regardless of what you're doing. I discovered this was the case when out of desperation I reinstalled the OS onto another SSD and since then my PC has been working flawlessly.

 

This however got me wondering, in both cases when the computer froze the HDD activity LED on the system would just stay lit, and I know that isn't really THAT uncommon, but here is my thought under the circumstances, whilst it was my SSD at fault, is it possible that the cache on a HDD failing would cause that sort of thing to occur? So a disk test which in this case was Seatools would just read the whole drive and not rely on the cache hardly at all, but day to day use would use the cache to keep the system running more quickly. To me it stands to reason that if the cache on the drive is failing, the drive would seem to freeze spontaneously at a fairly random time simply through using the data in the cache rather than reading the drive. Equally, if the drive locks up, that means it won't write data into the error log, could this also prevent a SMART event being recorded? Is it possible that SMART might not actually monitor the cache activity?

 

I know this all seems rather random, but at this stage, swapping out a HDD in someone's system seems more preferable than trying to RMA other components which as I've said, seem to check out alright.

[Senzelian]

- snip -

I recommend you to look for help in a different forum. You'll most likely not get an answer here.  :unsure:

[rattyocaster]

I recommend you to look for help in a different forum. You'll most likely not get an answer here.  :unsure:

Do you mean I should have put it in the help section? Or is it perhaps just a query which is a little too specific?

[Senzelian]

Do you mean I should have put it in the help section? Or is it perhaps just a query which is a little too specific?

Nah, I meant that LinusTechTips isn't exactly the right forum for such questions.

I think LTT is more a forum for beginners and just cool projects and other random stuff.

[scottyseng]

Recently I spoke with someone who said his computer kept freezing without any real warning and after doing a proper reboot the system would seem fine. I checked the error logs and came up with nothing and I ran hard disk and memory checks and they all checked out ok. However, at the time I was having similar issues with a PC of mine which had an SSD in it, then the Windows 10 upgrade came out and that pretty much killed my SSD to the point where the system will freeze every 5 minutes regardless of what you're doing. I discovered this was the case when out of desperation I reinstalled the OS onto another SSD and since then my PC has been working flawlessly.

 

This however got me wondering, in both cases when the computer froze the HDD activity LED on the system would just stay lit, and I know that isn't really THAT uncommon, but here is my thought under the circumstances, whilst it was my SSD at fault, is it possible that the cache on a HDD failing would cause that sort of thing to occur? So a disk test which in this case was Seatools would just read the whole drive and not rely on the cache hardly at all, but day to day use would use the cache to keep the system running more quickly. To me it stands to reason that if the cache on the drive is failing, the drive would seem to freeze spontaneously at a fairly random time simply through using the data in the cache rather than reading the drive. Equally, if the drive locks up, that means it won't write data into the error log, could this also prevent a SMART event being recorded? Is it possible that SMART might not actually monitor the cache activity?

 

I know this all seems rather random, but at this stage, swapping out a HDD in someone's system seems more preferable than trying to RMA other components which as I've said, seem to check out alright.

 

Just checking, but did you update the chipset drivers? And what are the specs of your PC? Have you tried just running the PC with just the SSD only?

 

Yeah, my cousin's PC kept blue screening (Samsung 840 Pro) after the upgrade until we updated the Intel Chipset Drivers and the Nvidia graphics driver.

[rattyocaster]

Just checking, but did you update the chipset drivers? And what are the specs of your PC? Have you tried just running the PC with just the SSD only?

Drivers and firmware are up to date. Actually the system that I'm currently struggling with is using a new 3.5" Seagate HDD (exact model escapes me though, supports SATA-600 apparently)as for the rest of the system, Asus A88XM-PLUS, 4GB HyperX Fury 1866MHz, AMD A6-6400 APU and that about covers it. PSU is just some bog basic 500w PSU which came with the case I know that isn't great but it should still be sufficient for this system checking e-mail and browsing the web.

[scottyseng]

Drivers and firmware are up to date. Actually the system that I'm currently struggling with is using a new 3.5" Seagate HDD (exact model escapes me though, supports SATA-600 apparently)as for the rest of the system, Asus A88XM-PLUS, 4GB HyperX Fury 1866MHz, AMD A6-6400 APU and that about covers it. PSU is just some bog basic 500w PSU which came with the case I know that isn't great but it should still be sufficient for this system checking e-mail and browsing the web.

 

Hmm, you might consider clean installing Windows 10. But I don't really see any reason as to why it's going bad for you.

[rattyocaster]

Hmm, you might consider clean installing Windows 10. But I don't really see any reason as to why it's going bad for you.

 

I'll have a look into reinstalling when I get chance. Strange thing is, if it is anything software related it must be an update that didn't install correctly rather than anything else as there's not really much installed onto it in the first place. It's why I'm wondering if it's HDD cache related, so writing to the drive and varied use seems ok, but repetitive tasks seem to cause it to freeze. It's just irritating that the freezes are difficult to replicate as it seems to be with different programs and different times which sort of makes me think it's not JUST software. I think I'm just going to have to have a play when I get chance :\ might try a reinstall but get a HDD on standby, they always come in handy anyway...

[Dabombinable]

@Captain_WD might have some knowledge.

[scottyseng]

I'll have a look into reinstalling when I get chance. Strange thing is, if it is anything software related it must be an update that didn't install correctly rather than anything else as there's not really much installed onto it in the first place. It's why I'm wondering if it's HDD cache related, so writing to the drive and varied use seems ok, but repetitive tasks seem to cause it to freeze. It's just irritating that the freezes are difficult to replicate as it seems to be with different programs and different times which sort of makes me think it's not JUST software. I think I'm just going to have to have a play when I get chance :\ might try a reinstall but get a HDD on standby, they always come in handy anyway...

 

Hmm, have you tried running an extended drive test, but yeah, usually those are signs of a dying hard drive or maybe bad system RAM? Usually the hard drive cache never goes bad / is too small to even make a difference in the system. My friend's laptop with a beginning to fail hard drive (Had no more sectors to relocate to / Caution in CrystalDIskInfo) would bluescreen randomly, from windows updates to just using the internet browser.

 

But I would try to reinstall windows, but do a full clean install (Use the ISO of the upgrade and load it onto a DVD or use a bootable USB) then completely delete the existing drive partitions using the install prompt. Since you already managed to upgrade the PC with Windows 10, you already have a key (Just keep skipping the install whenever you see a prompt for a key, and at the end, it will auto activate with internet).

[Dabombinable]

Another part of the HDD that may be causing issues is the controller board itself. With all of my old Maxtor HDD (notorious for having the control board die before any chance of mechanical failure-I can get new ones however) similar problems appeared before the HDD stopped working. For example at first I thought that my 420MB HDD had a mechanical failure due to the knocking sounds as the HDD became increasingly worse until the HDD light stayed on and nobody was home, however I was able to get it working again with an aforementioned control board replacement.

[Captain_WD]

Recently I spoke with someone who said his computer kept freezing without any real warning and after doing a proper reboot the system would seem fine. I checked the error logs and came up with nothing and I ran hard disk and memory checks and they all checked out ok. However, at the time I was having similar issues with a PC of mine which had an SSD in it, then the Windows 10 upgrade came out and that pretty much killed my SSD to the point where the system will freeze every 5 minutes regardless of what you're doing. I discovered this was the case when out of desperation I reinstalled the OS onto another SSD and since then my PC has been working flawlessly.

 

This however got me wondering, in both cases when the computer froze the HDD activity LED on the system would just stay lit, and I know that isn't really THAT uncommon, but here is my thought under the circumstances, whilst it was my SSD at fault, is it possible that the cache on a HDD failing would cause that sort of thing to occur? So a disk test which in this case was Seatools would just read the whole drive and not rely on the cache hardly at all, but day to day use would use the cache to keep the system running more quickly. To me it stands to reason that if the cache on the drive is failing, the drive would seem to freeze spontaneously at a fairly random time simply through using the data in the cache rather than reading the drive. Equally, if the drive locks up, that means it won't write data into the error log, could this also prevent a SMART event being recorded? Is it possible that SMART might not actually monitor the cache activity?

 

I know this all seems rather random, but at this stage, swapping out a HDD in someone's system seems more preferable than trying to RMA other components which as I've said, seem to check out alright.

 

Hey there rattyocaster, 

 

I would make sure the drive passes all test on the tool provided by the manufacturer (including the one for bad sectors). You can try other tools just to be on the safe side. Running chkdsk /r from the CMD is also an option to check if the drive is functioning properly. 

Does the disk usage hit 100% during those freezes? Can you hear the drive working intensively during the system freezes? I would try to run another drive on this system and see if the problem comes from the original one or from somewhere else.

 

Captain_WD.

[rattyocaster]

 

Hey there rattyocaster, 

 

I would make sure the drive passes all test on the tool provided by the manufacturer (including the one for bad sectors). You can try other tools just to be on the safe side. Running chkdsk /r from the CMD is also an option to check if the drive is functioning properly. 

Does the disk usage hit 100% during those freezes? Can you hear the drive working intensively during the system freezes? I would try to run another drive on this system and see if the problem comes from the original one or from somewhere else.

 

Captain_WD.

 

When it freezes the computer becomes completely unresponsive, the cursor moves but if you click on anything basically nothing happens, or if it does it's only stuff which is still in memory. Accessing task manager is practically impossible unless it was already open and even then sometimes switching tabs could be enough for task manager to stop responding as well, as for audible disk activity there's nothing at all other the hum of the motor. It really doesn't help that I don't currently have access to the computer to tinker with it, but if I can have some ideas for the next time I get to it that would be great. What I'm planning thus far the next time I get my hands on the system is running a full Seatools scan on it (Seagate drive, sorry Captain_WD) and a chkdsk /r in another system (to eliminate hardware issues in the initial host system) and at the same time I might have another drive which I can put into the system and install an OS onto, hopefully whilst I'm battering the initial drive with tests, installing and updating on another drive will hopefully show up whether the rest of the system has stability issues. I have already run memtest with 0 issues found so that is unlikely and the memory isn't overclocked  for the platform it's installed into either. Thanks for all your help and suggestions so far

[Captain_WD]

~snip~

 

Do post with the results of the tests. You might also run a full stress test to see if any of the other components have problems or are not functioning properly. Open the Resource Manager and leave it opened and if the system hangs again, check if and what is hitting 100% usage.

 

Captain_WD.