Question Raid 1 (mirroring) new HDDs + 100% active time. Pls help before crash...

Status
Not open for further replies.

JorgeIcarus

Honorable
Jan 17, 2015
17
0
10,510
Hello everyone.
A common issue, I know. And I have tried all the commonly suggested solutions.
I was wondering if the RAID 1 set-up I created could be the cause of this.
I have used the Win10 built-in wizard to create a Raid 1 (mirroring) 4 partition space.
Today I noticed some weird clicks and noises coming from my desktop.
The HDDs are brand new Seagate BarraCuda 4TB 3.5" SATA HDD/Hard Disk Drive ST4000DM004/DM004 .
I used to have the same problem with the previous data single HDD that eventually failed.
Avira does not detect any virus.
Check the screenshot below: Disk 0 is 100% active time reading at a speed of about 12 to 18 MB/sec. Disk 1 is at 8-10% active time with a matching writing speed of about 12 to 18 MB/sec.
The process that is causing this is System.
Any suggestions?



Capture.png
 
This has nothing to do with RAID. It's a system issue of high disk usage.

try this https://forums.tomshardware.com/faq/fix-100-disc-usage-by-system-service.2886539/
Unfortunately, that did not work.
Disk 0 active time is still at 100%.
It is strange that even though the speed at which Disk 0 is reading and Disk 1 is writing is exactly the same (Disk 0 and Disk 1 are mirrored, btw..shouldn't they write or read at the same time?), Disk 0 active time is 100%, while Disk 1 active time is at about 10%.
As I said in the opening post, I have tried all the common solutions to this problem, including editing the register and changing that 1 to 0 in the controller settings...nothing.

I have run a scan with MalwareBytes and I have got these results:
Capture.png


I believe that the ones related to William Hill Poker and uTorrent are probably false positives.
I am going to delete the Registry Key, but I don't believe that is what causing the 100% active time on one of my HDDs

Any other ideas?
 
Last edited:
When you say you used Win 10 built-in wizard, do you mean the one in disk management? If so, this is software RAID. Any reason no to use the RAID controller on your motherboard?

Raid 1 increases reliability at the cost of performance. Operations will be mirrored for each disk in the array, so you are constrained by the least performant of the bunch. If one of the disks is failing, your pc will be stuttering quite frequently until the disk fails completely. Also, Seagate aren't known for reliability and my past experiences with their disks give me no reason to dispute this (had about half of them fail prematurely).

My advice would be to look at SMART stats to see if there are any signs of a failing disk. Also, I would consider moving to hardware RAID.
 
When you say you used Win 10 built-in wizard, do you mean the one in disk management? If so, this is software RAID. Any reason no to use the RAID controller on your motherboard?

Raid 1 increases reliability at the cost of performance. Operations will be mirrored for each disk in the array, so you are constrained by the least performant of the bunch. If one of the disks is failing, your pc will be stuttering quite frequently until the disk fails completely. Also, Seagate aren't known for reliability and my past experiences with their disks give me no reason to dispute this (had about half of them fail prematurely).

My advice would be to look at SMART stats to see if there are any signs of a failing disk. Also, I would consider moving to hardware RAID.

Thank you for your reply.
The main reason why I chose to use software RAID is that I have read several articles that stated that Mainboards RAID controllers are actually software controllers. Moreover, in case of a controller failure, I should find the exact hardware replacement in order to access the HDDs data.
I agree about Seagate HDDs being not reliable, but these two are BRAND new ones! I bought them last month! And this issue with 100% disk time had already occured in the past with another HDD which eventually failed. I am keen on thinking that what actually made the HDD fail in the first place was the 100% active time for weeks and weeks...and this must be a Windows related problem!
Here are the SMART info for the 2 HDDs:
Capture.png

Capture1.png



----
So, quick update: now the active time of Disk 0 is down to 3-4%, while the Disk 1 skyrocketed at 100%! LOL
What I did was:
  • untick 'enable write-caching on device' for both HDDs of the Raid 1 array
  • click on 'reactivate disk' in the Disk Management windows (see image below). That apparently did nothing, though...as all the partitions on the disks still appear as 'resyncing'
Capture.png
 
Last edited:
So, I have apparently found out the reason why the Disks were alternatively 100% active time.
It is indeed the Windows Raid 1.
Supposedly, the Mirror Raid 1 needs to resync each time something such as a forced shutdown or a system error occurs.
The resynch will take, literally, AGES...over 5 days, if I am correct.
This is utterly STUPID and ridiculous.
More info here:
https://social.technet.microsoft.co...13cd5/dynamic-disks-raid-1-resynching-problem
https://serverfault.com/questions/107636/how-to-view-status-of-software-raid-1-resynching

The stupid thing is that it does not matter how large is the partition where the error occurred or how much data you have on it. The mirror rebuild will resynch the WHOLE disk, regardless.
In my case, we are talking about a 4 TB disk. Roughly 3700 GB or 3788800 MB.
Resynch is now taking place at 8MB/sec
That means: 473600 seconds...7893 minutes....131 hours...5 and a half days
And the resynch will have to restart if I reboot, according to what people wrote in one of those links above.

I need to find a different solution here. What if I cloned the disk? Is there any way I could just start mirroring without this lengthy process?
This is driving me mad...
 
Welcome to RAID. It's working exactly as intended. Why are you using RAID anyway? It's NOT a backup solution. It's intended for those applications where there can be no down time when a drive fails. Very unlikely that a home system fits this bill.
 
Last edited:
Backing up several steps....why a RAID 1?
What are you trying to protect against?
what do you mean by "several steps"? I am trying to protect against data loss.

Welcome to RAID. It's working exactly as intended. Why are you using RAID anyway? It's NOT a backup solution. It's intended for those applications where there can be no down time when a drive fails. Very unlikely that a home system fits this bill.
I am using RAID 1 as protection to avoid data loss caused by disk failure and data becoming inaccessible (redundancy error, is that what is called?). I thought this is the scope of Raid 1?

I considered RAID 1 as a sort of real time backup...I would keep swapping and replacing HDDs as they fail. Running a daily/weekly backup is just too much of a bother.

Unfortunately, I came to realize that Windows 10 is probably the cause of such extreme wear of my HDDs...all this 100% active time for no reason...surely that causes some wear of the discs?
 
"back up several steps" means...lets go back to the beginning, and discern the WHY of this RAID 1.

"I am trying to protect against data loss. "
RAID 1 is absolutely not that.

A RAID 1 only helps in the case of a physically failed drive, and you also need actual uninterrupted ops. For instance, if you were running a webstore, and downtime = lost sales.
Any business that runs their system in RAID 1 also has actual backups.

RAID 1 does nothing for the more common forms of data loss. Accidental deletion, virus, ransomware, file corruption, etc, etc, etc.

In this case, the RAID 1 is doing exactly what it is supposed to be doing, fixing a corrupted RAID array.
A force shutdown leaves it in a corrupted state. The two physical drives are not in sync. When you power it back up, it needs to reconcile both drives. Taking hours or days, the larger the drives get.
Exactly what you're seeing now.


You need an actual backup routine, rather than a RAID 1.
And properly done backups are far more versatile than a RAID 1.

Read here to get started down that path:
 
what do you mean by "several steps"? I am trying to protect against data loss.


I am using RAID 1 as protection to avoid data loss caused by disk failure and data becoming inaccessible (redundancy error, is that what is called?). I thought this is the scope of Raid 1?

I considered RAID 1 as a sort of real time backup...I would keep swapping and replacing HDDs as they fail. Running a daily/weekly backup is just too much of a bother.

Unfortunately, I came to realize that Windows 10 is probably the cause of such extreme wear of my HDDs...all this 100% active time for no reason...surely that causes some wear of the discs?

What they are trying to say is that Raid 1 is not a backup solution. What happens if both drives die? Or you get the crypto virus? Or another virus that makes your system unbootable? You'd just be copying that virus to both drives... You are better off going back to a normal disk setup without Raid. Getting an external disk and using any free backup program to run daily backups (I personally would suggest Veeam Agent for Windows which is free).

And yes. Your Raid is working as intended. The drives are continuously mirroring each other so data is going to be going back and forth between the drives to update each one with correct data.

Also on a software raid if your OS gets corrupted and you cant boot then your raid is broken... this is why on board or raid cards are important.

I would just break the raid and make backups. That raid in the long term will just cause more problems then it is worth.
 
"back up several steps" means...lets go back to the beginning, and discern the WHY of this RAID 1.
hehe! it's the 'back up' bit that got me...I thought you were talking about a new type of backup...several steps backup! LOL

Read here to get started down that path:
Cool. thanks.

I would just break the raid and make backups. That raid in the long term will just cause more problems then it is worth.
Yup. Figured that out. Cloning the drive right now...happy days.

It's not Windows 10. It is the RAID array, trying to rebuild itself.
I was talking about the 100% time issue which is so common with Windows 10 and that I have encountered before...not in this instance.
 
Thank you for your reply.
The main reason why I chose to use software RAID is that I have read several articles that stated that Mainboards RAID controllers are actually software controllers. Moreover, in case of a controller failure, I should find the exact hardware replacement in order to access the HDDs data.

Well, I agree integrated RAID controllers aren't the best solution, but my foray into Windows' software RAID and Storage Spaces wasn't the best, so I ended up buying a proper RAID controller (you can get decent server-pulled raid cards for some 50 bucks, battery included).

It's not like software RAID is a problem, it works and you can move the disks to a similar OS, which is nice. What killed it for me was that it had a tendency to lose sync whenever there was a system crash. Trusting a storage solution on windows not crashing was a bit too much for me.

I agree about Seagate HDDs being not reliable, but these two are BRAND new ones! I bought them last month! And this issue with 100% disk time had already occured in the past with another HDD which eventually failed. I am keen on thinking that what actually made the HDD fail in the first place was the 100% active time for weeks and weeks...and this must be a Windows related problem!

Disk failure rates are actually pretty high during the first months of use. If it survives the first year, it will probably last another two or more depending on model and usage. Makes sense if you think about it, early failures due to constructive defects and late failures due to wear.

It seems you are running a pair of disks with 4 volumes, and mirroring the volumes in Windows. I'm not sure how windows handles this, but I wouldn't be surprised if it tries to mirror everything simultaneously. If it does, you can imagine how reading / writing from four different partitions in the same disk will be awfully inefficient. I'm curious as to what kind of queue depths you are getting while synching.

If I were to do software RAID, I'd probably stick to a single partition and separate whatever I wanted into VHDs, which I'd then mount as drives. You might want to take a look at storage spaces too, as it seems to be the way Microsoft is going forward with Windows storage (the article for disk management now links to storage spaces). Here are two articles in case you want to read a bit into this.

https://hetmanrecovery.com/recovery...e-or-mirrored-volume-in-windows-7-8-or-10.htm
https://betanews.com/2014/01/15/windows-storage-spaces-and-refs-is-it-time-to-ditch-raid-for-good/
 
Last edited:
Status
Not open for further replies.