Reinitialize disk surface on Raid JBOD Controllers

Experiences with hard disks, SSDs, USB devices, hard disk controllers, motherboards and so.

Reinitialize disk surface on Raid JBOD Controllers

Postby BigXor » 2014.10.30. 19:23

I have used the HDS disk utilities on the following cards in Raid 0 Mode. Setup in JBOD mode slows the tests down to a crawl.

LSI9211-8i
Dell H710
Dell H810

For Reinitialize Disk Surface:

When I create a pass-through disk and start a Reinitialize disk surface test the performance is ~9MBs on the Raid cards. When I create a single drive Raid 0 disk with no read ahead and write-through the performance is ~65MBs.

Motherboard Intel AHCI SATA ports is ~50MBs


My question is: does the software correctly detect and remap sectors using a single drive Raid 0 ??

The MB Intel controller I know works because I have 1 drive out of 4 testing that has shown 10 bad sectors so far.

I have 15 drives on a Dell H810 (LSI SAS2208 Controller) connected through a Chenbro CK23601 36 port expander that needs to be tested for bad sectors. (HDS shows SMART on both Raid and JBOD on all the controllers).

So the final question is: Will HDS detect and remap sectors in single drive Raid 0 since it is the fastest setup?

JBOD is out of the question at 9MBs. Testing nine 3TB and six 4TB drives would take weeks. I know for fact that the MB controller works correctly, but I only have 4 ports available. I want to test all the drives at once and be done with it.
BigXor
 
Posts: 2
Joined: 2014.10.30. 18:52

Re: Reinitialize disk surface on Raid JBOD Controllers

Postby hdsentinel » 2014.11.03. 08:57

Thanks for the information and the experiences !

Can you please tell me the hard disk model(s) as well? Or (even better) use Report menu -> Send test report to developer option?
This way it is possible to check the situation, verify if there is any kind of compatibility issue between these controllers and the appropriate hard disk models, which may caused a synchronization issue and reduced performance. Maybe a driver and/or firmware upgrade of the controller can help in this situation.
While this is not really related to Hard Disk Sentinel, would be interesting to check and even consult with LSI / Dell.


> My question is: does the software correctly detect and remap sectors using a single drive Raid 0 ??

Yes of course.
As even in this mode, it is possible to access the complete surface area (minus some sectors reserved by the controller where the administrative area of the RAID configuration located), then it is possible to reveal and fix possible issues even in single drive RAID 0 configuration.


> So the final question is: Will HDS detect and remap sectors in single drive Raid 0 since it is the fastest setup?

Yes, I can confirm that.
User avatar
hdsentinel
Site Admin
 
Posts: 2334
Joined: 2008.07.27. 17:00
Location: Hungary

Re: Reinitialize disk surface on Raid JBOD Controllers

Postby bash » 2016.01.07. 05:58

Sorry for thread necro but this pretty much captures the spirit of my question.

I have a IBM M1015 flashed into IT mode which is just passthrough this connects to a Intel-RES2SV240 which connects to a Norco 4224 SAS backplate in its 24U enclosure.
This is a pretty popular setup for both unRaid and ZFS for JBOD.

Now the question I have 3 Toshiba Enterprise SSD 3b/s drives that performed 100mb/sec+ on read surface tests and found bad sectors.

I then started the Reinitialize disk surface test but the speed is incredibly slow. Even just a 400GB SDD will take 90+ hours if the estimate holds.

All three drives are showing 1.37mb/s or 1.38mb/s speed for the Reinitialize test. Any Idea's?
bash
 
Posts: 2
Joined: 2016.01.07. 05:54

Re: Reinitialize disk surface on Raid JBOD Controllers

Postby hdsentinel » 2016.01.07. 15:47

Thanks for your message and thanks for sending the developer report.

Did you try a simple write test, just to check the actual write speed of that SSD?

Generally, the Reinitialise Disk Surface test performs multiple pass overwrites before clearing the data (which is also a write operation) and then read back to verify the sector contents.
So usually if the Reinitialise Disk Surface test is slow, it means that the write speed of the device may be slow.

But this seems really slow, so I'm not really sure if the issue is related to the
- SSD itself
- the disk controller itself which may have some compatibility issue with this device.

Personally, I'd try to connect one of these SSDs completely differently, for example to the integrated Intel chipset controller, just to verify if there is any difference: there is any improved speed in both the "simple" Disk menu -> Surface test -> Write test and also the Disk menu -> Surface test -> Reinitialise disk surface test.
User avatar
hdsentinel
Site Admin
 
Posts: 2334
Joined: 2008.07.27. 17:00
Location: Hungary


Return to Supported Hardware

Who is online

Users browsing this forum: No registered users