Bug Recognizing JMicron H/W RAID0

Post here if you encounter any problems or bugs with the software.
voyager62
Posts: 24
Joined: 2014.10.16. 05:55

Bug Recognizing JMicron H/W RAID0

Post by voyager62 »

After I installed a dual mSATA to SATA RAID card in my laptop and configured it for RAID0, I noticed that HDS doesn't properly recognize it as a single drive. The size is correct for the combined drive, but since it sees it as 2 separate drives the total size is doubled and drive numbers are now different from what Windows reports. Even if I hide one of the SSD's in the main window, the drive numbers aren't corrected. This is another reason it would be nice to be able to sort by drive letter. What use is the drive number if it isn't the same as what Windows reports? Also the bytes read/written that HDS reports is off by a factor of 10^6 as far as I can tell. There is no possibility of these SSD's writing 5 Exabytes in their lifetimes let alone 17 days.
HDS RAID0
HDS RAID0
HD Sentinel RAID0.png (80.77 KiB) Viewed 7075 times
Windows RAID0
Windows RAID0
Disk Management RAID0.png (38.55 KiB) Viewed 7075 times
User avatar
hdsentinel
Site Admin
Posts: 3115
Joined: 2008.07.27. 17:00
Location: Hungary
Contact:

Re: Bug Recognizing JMicron H/W RAID0

Post by hdsentinel »

Thanks for your message and excuse me for the possible troubles.

However, I can confirm that no, absolutely no bug here, everything working exactly as should.

> I noticed that HDS doesn't properly recognize it as a single drive.

Yes of course. This is exactly the purpose of the RAID detection: to detect all disks, the members of the RAID array(s) configured.
This is why Hard Disk Sentinel does not show as single drive, but exactly detects and displays all (currently two) devices, as they together form the array.

> The size is correct for the combined drive, but since it sees it as 2 separate drives the total size is doubled

Yes of course. This is also completely normal: as the size suggests the total usable capacity of the complete array, from the viewpoint of the OS.
Otherwise users may be confused to see that (for example) a total of 500 GB partitions reside on smaller disk drive...
Sorry for the confusion, but this is normal.

> and drive numbers are now different from what Windows reports.

Yes of course, this is also completely normal in RAID configuration.
Generally Hard Disk Sentinel always follow the disk orders of Windows Device Management, exactly to allow identification of disks better.
However, when RAID array found, it is completely normal that all members of the array enumerated and displayed - before following the order of other disks as they follow in Windows Disk Management.

There would be no point in other solutions, for example showing only one disk there - and the other after others - this would be really confusing, as (if we speak about a RAID array) we may expect that all members should follow each other.

> Even if I hide one of the SSD's in the main window, the drive numbers aren't corrected.

Yes of course. Hiding one of the SSDs from the main window does not affect disk numbering: if you hide a device, then it is still detected and monitored (so a drive number is assigned and used by Hard Disk Sentinel to address it).

> What use is the drive number if it isn't the same as what Windows reports?

Windows does not detect the RAID members - but I do not see it is "bug" or limitation of Hard Disk Sentinel, but exactly the opposite.

If you prefer, you can delete the files DETJM.DLL and DETJM2.DLL from the folder of Hard Disk Sentinel.
Then it will not attempt to detect this RAID configuration - and will show only the combined RAID array, as the virtual device as drive #0, without ANY status information, without enumerating the RAID members and their status information.


> Also the bytes read/written that HDS reports is off by a factor of 10^6 as far as I can tell.
> There is no possibility of these SSD's writing 5 Exabytes in their lifetimes let alone 17 days.

This seems so interesting: the SSD and its current firmware may report the total lifetime writes differently than previous tested models.
Can you please use Report menu -> Send test report to developer option?
Then I can check the actual status, the total amount of written data and correct this.
Excuse me for that, yes, this needs to be corrected of course, thanks for your attention.
voyager62
Posts: 24
Joined: 2014.10.16. 05:55

Re: Bug Recognizing JMicron H/W RAID0

Post by voyager62 »

If there isn't a bug recognizing JMicron RAID0, why when I set the array up and transferred the system over USB, did it show up properly even though HDS see the individual drives over USB using the bridge chip I'm using? You can see it how it appears here when I tested it with other SSD's I had. Shouldn't the display be consistent no matter which RAID mode the board is configured for? This is why it was a surprise that the display was different after installing it into the laptop.
This seems so interesting: the SSD and its current firmware may report the total lifetime writes differently than previous tested models.
I found a firmware update for the Sandisk X300 SSD's on Dell's website, so I'll send a test report, then update the firmware and see if there's a difference in the reported bytes written.
RAID0 on USB
RAID0 on USB
HD Sentinel RAID0 USB.png (64.04 KiB) Viewed 7072 times
RAID1 on USB
RAID1 on USB
HD Sentinel RAID1 USB.png (75.17 KiB) Viewed 7072 times
SPAN on USB
SPAN on USB
HD Sentinel SPAN USB.png (83.42 KiB) Viewed 7072 times
JBOD isn't properly recognized by Windows or isn't supported by the SATA-USB Bridge.
JBOD on USB
JBOD on USB
HD Sentinel JBOD USB.png (84.91 KiB) Viewed 7072 times
User avatar
hdsentinel
Site Admin
Posts: 3115
Joined: 2008.07.27. 17:00
Location: Hungary
Contact:

Re: Bug Recognizing JMicron H/W RAID0

Post by hdsentinel »

> Shouldn't the display be consistent no matter which RAID mode the board is configured for?

Yes, the display should be consistent of course.
But there may be situations which can affect the detection: USB controllers, USB/ATA converters (especially if used with external RAID devices) may affect how the devices can be accessed and listed.
According the experiences, such RAID devices may work completely differently over USB and SATA connection.

Generally the expected display is that the RAID members should be displayed - and the RAID device itself (which has no additional information, no additional status information) should be NOT displayed (but in some cases, it may be displayed too).

This is independent from the RAID level of course.

But if USB enclosure / adapter used, it is possible that it may not comatible with some modes of the RAID controller. JBOD / Clean mode (when the drives in standalone configuration) usually not supported by most USB adapters as they may not able to address both drives (even if they may be identified and their status can be detected in Hard Disk Sentinel).

Also during creation / configuration / setup of RAID arrays, it is possible that some time or a power cycle may be required in order to follow the change and Hard Disk Sentinel to display the changed configuration.
In worst case, it is possible that disconnection/reconnection of the external RAID device and/or complete restart of the software may be required in order to see the status information after the initial configuration.

Usually the best is to please use Report menu -> Send test report to developer option, may help more than images.
This way I can check the actual situation: examine the drive configuration of Windows Disk Management, verify what is the "raw" response of the RAID controller and the external USB/ATA converter.
This may give information to verify what can be wrong and how things can be improved - if can and should be improved.
Generally I'm not sure if there is any problem with Hard Disk Sentinel, it attempts to identify the drives, detect and display the status information in all possible ways - if the RAID chip and the USB bridge chip both allows that.
Post Reply