100% HDD Usage

The one I bought exclusively for surveillance is the Seagate ST8000DM004 (8TB). That's the one that is causing me issues. I switched to a Hitachi HDS5C3020ALA632 (2TB) which causes no problems but is pretty small (it's the only empty drive I had).

@IAmATeaf I did run the Windows 10 diagnostics on that drive before posting here. It showed no problems. After reading your post I downloaded the Seagate tool. The quick test showed no problems. I am running the "Long Self Test Running" test but it takes forever.
 
The "Long Self Test" finished and it says it passed. I have to say it's kind of strange that the HD goes to 100% and slows down my entire system with BI writing less than 10MBytes/s. But since switching to my 2TB HDD everything works fine. Just for testing I switched back to the 8TB drive and same thing. After a while it goes to 100% and slows down the system.
 
  • Like
Reactions: samplenhold
I deleted 512GB of data from my new 8TB drive (filled 90% before deleting 512GB) and then started to copy 512GB of BI data (each file 3GB) from the old 2TB drive. I did this with BI shut down so it doesn't write to the old drive while I am copying data. Initially the speed was fluctuating between 10-100MBytes/s (according to task manager). Active time in task manager was close to 100% (average response time up to 1s) which is understandable considering the relatively high copy rate. But the average transfer rate was still fast. After about 30 minutes the average response time increased (fluctuating a lot between some 500ms to 10s). Write speed decreased to some 1MB/s to 5MB/s). The system response time decreased significantly (especially loading webpages). That matches what I experienced with BI. i.e. when I start BI for the first 20-60 minutes things are ok but then the system slows down. Also initially when I started they copying process Windows told me it will take about 1h. After those 30 minutes it told me it will take 10 hours, i.e. 10x longer.
 
  • Wow
Reactions: samplenhold
Sounds like a faulty drive to me, the issue you have is that you won’t be able to RMA it unless their tools report an error.

What I would do is fit the drive in another system and keep running the long test on it hoping that it eventually errors?

Or in fact it might be worth just testing copy speeds in another system to see if it behaves the same.
 
  • Like
Reactions: rfj
I am running more of the extended tests but I am not confident it will find anything. I was planning to use this drive for some other purposes/data but now I am not sure that's a wise thing to do....
 
  • Like
Reactions: looney2ns
You could store backup images of your C:\ drive, and use it as an external USB drive if you get a shell/case with USB 3.0,,,,,
 
  • Like
Reactions: looney2ns
From memory there is an known Win 10 bug that causes 100% disc usage after Windows Update. If it does it without BI running I suggest using DISM commands to repair windows.
 
Last edited:
I contacted Seagate support and they offered to replace my drive. I got it today and put it in the system. At this point it is filled 1.5%. I stepped out of the house for an hour and when I got back to my computer websites wouldn't pull up and the system was sluggish. Turns out the new disc has 100% activity time. I only write like 10MBytes/s. Interestingly, I shut down BI about a 1minute ago and it's still at 100%. Looking at the Resource Monitor System is still writing stuff to the disc. It appears as if Windows did cash the files and just slowly write them. Malwarebytes was shut down and Virus & threat protection is off as well.
 
Memtest 86?

If it's caching to disc I wonder if the memory is ok. Typically you get caching when Windows runs out of memory.

If that's ok, run DISM commands on your system to check Windows for errors.

BTW you need to run Memtest for a long period ie as much as 24 hrs.
 
For now I switched back to my old 2TB Hitachi drive. No problems... I think that Seagate drive just can't handle 10x 1MByte/s streams for a total of 10MBytes/s

I also did a quick test copying a 3GB file from the Hitachi Drive to the Seagate drive and from the Seagate drive to the Hitachi drive. The former took 85s whereas the latter took 25s. Granted, the Hitachi is 7200rpm and the Seagate 5400rpm. In any case, that results in about 35MBytes/s for the Seagate. That's with one large file. So I can see that it starts having problems with 10x 1MByte/s = 10MBytes/s streams.

I also ran CrystalDiskMark on both drives. The Seagate is a lot faster for sequential reads (almost 3x, about the same for random reads). The Seagate can keep up with sequential writes when only one queue is involved. However, it gets completely trashed for sequential writes with 8 queues (scenario for BI) and for all random writes. I wonder if all SMR drives are like this.
 

Attachments

  • Hitachi_2TB_7200rpm_CMR.png
    Hitachi_2TB_7200rpm_CMR.png
    19.9 KB · Views: 3
  • Seagate_8TB_5400rpm_SMR.png
    Seagate_8TB_5400rpm_SMR.png
    20.4 KB · Views: 3
Last edited: