Running Tests Simultaneously

[Sibelius]

I have just bought 2 new external 18TB HDs and am running all 4 recommended tests for a new drive before I put them in my NAS. A few questions:

1. While I'm running an Extended test on both drives, is it possible to simultaneously run a Surface Read test in parallel? HD Sentinel seems to allow it in the menus at least (the disks are available to be selected). But I don't know if selecting it will simply mean the Read test task will be queued in HD Sentinel and automatically started after the Extended test is over
2. I was wondering what is the purpose of running the Read test before running the Reinitialize Surface test. I would have thought the latter would cover all the bases.
3. For the surface test, is there a particular recommended configuration to be used (Random test, Butterfly test, etc) to ensure extra robustness?
4. As long as the disk temperature doesn't exceed 43C, is that considered acceptable or does the temperature need to be lower? I'm thinking of setting up a fan to blow on the external enclosures during the tests.

[hdsentinel]

> 1. While I'm running an Extended test on both drives, is it possible to simultaneously run a Surface Read test in parallel?

Yes it is possible - but it is not really good idea.

> HD Sentinel seems to allow it in the menus at least (the disks are available to be selected).
> But I don't know if selecting it will simply mean the Read test task will be queued in HD Sentinel and automatically started after the Extended test is over

No, the Read test will not be queued, it starts immediately when you select it.
The Extended Self Test runs "inside" the hard disk drive, when the disk does not perform any actions, no read/write commands performed. The execution of that Extended Self Test is completely automatic, we have no control about how it runs and we can only see the progress (in 10% steps) and the result (OK or problem).

You can start a Disk menu -> Surface test -> Read test but then the Extended Self Test will be paused by the disk drive automatically and will only continue when the Surface test finished.
So the hard disk drive will show eg. 90% remaining for many hours.


> 2. I was wondering what is the purpose of running the Read test before running the Reinitialize Surface test. I would have thought the latter would cover all the bases.

Yes, you're correct. If you plan to perform a Reinitialise Disk Surface test and also if you used the Extended Self Test already - then there may be no reason to use the Read test.
The Read test may be interesting if you prefer to view the performance curve of the hard disk drive and to reveal possible slower areas. These are not detected/reported by the Extended Self Test.
Please refer to the Help which gives additional details about the tests, their type and difference between them:

https://www.hdsentinel.com/help/en/62_testfaq.html
https://www.hdsentinel.com/help/en/58_test.html
https://www.hdsentinel.com/help/en/61_surfacetest.html


> 3. For the surface test, is there a particular recommended configuration to be used (Random test, Butterfly test, etc) to ensure extra robustness?

The default setting (sequential scan) is perfect for most situations. The other options may be interesting in special cases only, for example if the beginning of the drive has problems and we want to check if other areas (eg. the end of the drive) is better and the disk drive is still usable with smaller capacity.
You can use the Random / Butterfly option if you want extra "stress" for the drive, but you can leave the setting on default.


> 4. As long as the disk temperature doesn't exceed 43C, is that considered acceptable or does the temperature need to be lower?
> I'm thinking of setting up a fan to blow on the external enclosures during the tests.

If the temperature does not increase further even during the test, then it is absolutely perfect. Usually the temperature can rise with 10-15 Celsius (or even more) during an intensive test, so if 43 would be the idle temperature then I'd surely recommend an extra fan - but if you see 43 Celsius during the test, then it is perfect.

[Sibelius]

Thanks for the detailed response. Based on what you said here and also in this other thread (https://www.hdsentinel.com/forum/viewto ... 32&t=13493), I think it would be helpful if you modified the FAQ to mentioned that an Extended Test is redundant and that the Read Surface test will cover it and more. I understand what you're saying about the cables playing a role in the Read test, but it seems that the recommendation would be to run an Extended test only if a Read test fails, which seems to be all-inclusive. Correct me if I'm mistaken.

I am dealing with an issue now where the Extended Test seems to be hanging. It was supposed to take 33 hours, but when 33 hours passed, Sentinel simply keeps resetting the Elapsed time to around 32:33 and keeps saying remaining time is 1 minute. This has been going on for over 12 hours now (total 45hrs+ to test two 18TB drives simultaneously [USB3 connection]). I'm not sure what to do--do I just wait more or does this mean there is an issue with the software and that I should stop the test? No issues found so far, but I don't know if I can trust it if I have to terminate it manually.

Also, I am still a bit sketchy on whether Reinitialize surface test already does everything that Read surface test does. Could you clarify this? I'm fine taking several days doing tests since these drives are going into my NAS for years of service and I want to be absolutely sure everything is in perfect condition. But I don't want to waste time on redundant tests.

[Sibelius]

I'm somewhat frustrated right now. After days of running an Extended test and being stuck at 32 hr mark as I asked yesterday, this morning I saw an "Unknown" status on the disks being tested:


(image sharing doesn't seem to be working. Here is the location of the image: https://ibb.co/L074yt9)

I've seen this behavior before with HD Sentinel with my main SSD (SATA connection inside the computer) suddenly becoming "unknown". I exited HD Sentinel and reopened it, and now see this:


(image sharing doesn't seem to be working. Here is the location of the image: https://ibb.co/4d0zP6s)

It essentially says the Temperature has reached 57C on this drive(!), whereas during the entire test it was showing 36C, so I did not turn on the fan, not realizing that the damn thing was hanging. Now I'm wondering if I have damaged or shortened the lifespan of two brand new drives. Can you please advise on this?

I'm back to square one since I have no idea whether the Extended test was successful after 3 days on this test...

[hdsentinel]

> I am dealing with an issue now where the Extended Test seems to be hanging.

Generally I can confirm that it is "normal". As I tried to explain, when the Extended self test runs inside the hard disk drive, it is completely managed by the drive and we can only see a progress in 10% steps and the result (OK or failure). The manufacturer provides an estimated time for the test - but it is applicable only if the hard disk drive is completely idle (no other reads/writes by possible other software and/or Windows which could delay the test).

If the test delayed because of some background activity, then it seems "stuck" and waiting for completion. This can be completed any time as the duration of the test already finished (this is why "1 minute" displayed).

Please check this page:
https://www.hdsentinel.com/kb/category/ ... -back.html
as it describes the situation in general.

> I'm not sure what to do--do I just wait more or does this mean there is an issue with the software

Absolutely not.
Hard Disk Sentinel started the test and waiting for its completion - which can happen any time, but as the test is running by the hard disk firmware, we have no control over it.

Sometimes this is why the Surface test -> Read test may be better, as then we can see precise control about the progress and we can see the results during the test.
Plus we can stop and restart from specific position as described at

https://www.hdsentinel.com/kb/category/ ... drive.html


> Also, I am still a bit sketchy on whether Reinitialize surface test already does everything that Read surface test does. Could you clarify this?

The Read test performs a complete surface scan, so reads all sectors.
The Reinitialise Disk Surface test performs multiple overwrites with different patterns, then clears the sectors and reads back to verify the results.

So generally the "read" command (used by the Reat test) is included in the Reinitialise Disk Surface test too (and the previously mentioned Write+Read test too, which may be faster).

If you open Disk menu -> Surface test, in the window you can check the details/descriptions of each tests, this may give additional details.

[hdsentinel]

This seems so weird, like something could block accessing the hard disk status/details at all.
Is it possible that you used the option at Configuration -> Disk Control -> Disable detection of USB and SCSI drives ?
That option designed to disable detection codes and can result in such "unknown" state for the drives (depending on chipset and drivers, maybe for SATA drives too).

I received the report about the configuration, checked and based on that, I see (as displayed on the Information page):

Disk Controller . . . . . . . . . . . . . . . . : Standard SATA AHCI Controller (AHCI) [VEN: 8086, DEV: 8C82] Version: 10.0.19041.1288, 6-21-2006

which means that the disk controller runs with Standard AHCI controller driver.

In most cases, there are no problems with this solution (in some cases, this is even better than the driver provided by the manufacturer) but with some controllers/chipsets, this standard driver can cause issues (degraded performance, incorrect or partially missing status information reported and so).
So I'd recommend to try installing the proper Intel chipset driver for the system, just to be sure.

Also if you'll ever see similar missing or unknown status information, please use Report menu -> Send test report to developer option (or manually create + save the test report at Configuration -> Send tet report page) without restarting anything as it can help to determine what happens, what may cause blocking of proper operation.


> It essentially says the Temperature has reached 57C on this drive(!), whereas during the entire test it was showing 36C,
> so I did not turn on the fan, not realizing that the damn thing was hanging.

This sounds so interesting.
There is good chance that the internal power management function of the USB enclosure caused troubles. Some of such USB-adapters may not "notice" that the hard disk is running the extended self test (really) and can happily act as if the drive would be sleeping. So even during the detection, the previous temperature value may appear and/or even provide unknown status.
This can also confirm why the disk test progress did not change too.

I'd surely use the Disk menu -> Surface test -> Read test in this case instead of the Extended self test on this drive, exactly because it will surely show the status/progress better and you can see the disk temperature changes. If there seems no change for very long time (during any test) then we can suspect that there is something not correct: the disk controller/chipset or the USB adapter can cause issues in reading the proper temperature value (which should go up of course during the test).


> Now I'm wondering if I have damaged or shortened the lifespan of two brand new drives. Can you please advise on this?

No, I can confirm that it the disk drive surely NOT damaged, it should able to tolerate this temperature for a relatively short period: the hard disk not used on constant high temperatures for very long time and it had the opportunity to properly cool down.


> I'm back to square one since I have no idea whether the Extended test was successful after 3 days on this test..

I'm afraid the Extended self test may really not gave the proper results in this case. Probably the USB adapter and/or the chipset drivers or the drive firmware, but something caused confusion.
Sorry for that, I'll examine the report specifically and check if there is something to do (eg. a warning message in similar situations when the Extended self test selected or so).

Personally I'd use the Disk menu -> Surface test -> Read test ONLY for now.
Yes, considering the excessive long time (and the fact that the USB adapter/enclosure may be not 100% cooperative) I'd not use the Reinitialise Disk Surface test (or the Write+Read test) for now, unless the above mentioned Read test shows an error.

[Sibelius]

Unfortunately, I have no idea how long the HDs were stuck in 50C+ temperatures. It could have been short or very long. During the test, while the Sentinel was still connected, the temperature reported was a very suspicious flat line at exactly 36C, so I'm assuming that temperature was wrong. Definite concerns about the HDs being affected at this point.

I ran the Read test and it completed successfully within 27.5 hours for both disks. However, I doubt skipping a write test is a good idea, especially given the issues that occurred. So I'm running the Write + Read test currently.


Based on everything that you said about the tests, I would once again suggest modifying the FAQs. The only tests needed should be:

1. Short-Self Test
2. Reinitialize Disk Surface

This is because the Extended test is already a subset of the Read test. Seems like Extended can be run only if a Read test produces errors, just to verify the issues weren't caused by cable connections. The Extended test does have it's dangers though, as my case showed here.

And then the Reinitialize test also already includes a read test after a write test it does. As such, the Read test itself is redundant. Finally, I understand your point about using the Write+Read test instead of Reinitialize to avoid multiple rewrites and save time. Nevertheless, just by skipping the extra redundant tests saves a lot of time and I think it makes sense to just use the Reinitialize test. If I missed something, please let me know.

[Sibelius]

One more quick question: As I'm running the Write+Read test, the Write test has finished and it's now running the Read test. I don't see the data for the Write test anymore. Is there a way to retrieve the Write test results once the test is completer or does HD Setinel overwrite all results with the last test run? I want to save both the image for the blocks and also the CSV export of the write speeds and I hope it is saved somewhere in HD Sentinel.

[hdsentinel]

> This is because the Extended test is already a subset of the Read test.

Not really: if we want to be precise, exactly the opposite: the Exteded Self Test performs a quick internal check (similarly to the Short Self Test) plus makes a surface scan, similarly to the Read test (just the Extended Self Test stops on first error and have some issues as discussed above).
The Read test detects/reports possible issues in a detailed way - as it shows possible slower areas, retries, can be paused or restarted from specific position (plus does not stop on the first possible error).

I'd say if we already used the Short Self Test plus the Disk menu -> Surface test -> Read test, then it is not required to use the Extended Self Test.

> And then the Reinitialize test also already includes a read test after a write test it does. As such, the Read test itself is redundant.

If the Extended self test is not started - I'd surely recommend using the Read test. While you feel it is redundant as the read command is included the Reinitialise Disk Surface test, it may be still useful to reveal possible issues (eg. quickly reveal of an external USB hard disk connected to USB 2.0 slot / USB 2.0 hub which can reduce the performance) plus a read scan is safer than performing a write test immediately.

I just added the following to the F.A.Q. section:

https://www.hdsentinel.com/faq.php#tests

"To save time (especially on a high capacity hard disk eg. 8 TB or above) step 2 (Extended self test) may be not required and instead of step 4 (the Reinitialize disk surface test), the Write+Read test can be used to reveal and stabilize possible problems."

[hdsentinel]

I'm afraid the image related specifically for the write pass not saved: when the Read test starts in the Write+Read test, the disk surface map cleared (as you can see) and the result will show the combined surface map.

However, you can export the measured write and read speeds of each blocks: just right click on the surface map and select "Export transfer speed values" option to export them to CSV file.

The CSV file will contain (for example):

BLK,W (MByte/s),R (MByte/s),RES
0,100.28,112.87,OK
1,117.98,114.35,OK
2,109.73,116.07,OK
3,121.21,114.84,OK
4,107.58,112.82,OK
5,117.33,113.70,OK
6,116.59,114.58,OK
7,117.10,117.18,OK
8,116.99,116.10,OK

and so, to indicate both the Write and Read speed of each blocks.