0

I am building a NAS/home server, so I bought new WD Red Pro hard drives (18 TB), and the store didn't package them very well during shipment, so I decided to test the Hard Drives for bad sectors. They are empty, so data recovery wasn't important, but the speed of the test was - I wanted the fastest way to mark bad sectors/blocks of the hard drive.

This answer recommended Easeus Partition Manager with its Surface Test feature as the fastest way to mark bad blocks/sectors without recovery.

Before I ran the surface test, I tested the speed of my disks with Crystal Mark: around 200 MB/s for continuous write and around the same for continuous read.

When I started the surface test, Easeus Partition Manager displayed 10,000-12,000 MB/minute as the surface test speed. Now, the confusing part for me is that I was under the impression that the process of surface test is a 2-step process:

  1. Attempting to write to the disk sector
  2. Attempting to read the written data from that sector

12000 MB/minute of the test is the same as 200 MB/s when converted to the per second value. This should be exactly the speed of write or read alone, but not both. Isn't it the case that having to both read and write should yield half of that surface test speed = 6000 MB/s? And if so, does it mean that Easeus Partition Manager only writes or only reads, instead of doing both, to surface test a hard drive, and then how reliable is their test process? I.e., if the test says (after completing in 30 hours) that I have 0 bad blocks, do I really have no bad blocks?

P.S. I connected these internal drives externally and ran the test using a SATA to USB 3.0 adapter--if that matters.

InfiniteLoop
  • 121
  • 7
  • 1
    You do know you will see significant speed performance drops, due to the speed differences, between SATA and USB 3.2 Gen 1? – Ramhound Jun 13 '22 at 19:31
  • That's not the question. I am not claiming that something is too slow and asking "why". I am asking why the surface test seems to be TOO FAST. It doesn't seem to spend enough time on each sector to BOTH write to and read from it. Is it doing just one of those operations as part of the surface test process? And if so, is that really enough? – InfiniteLoop Jun 13 '22 at 19:35
  • 2
    I am not sure anyone outside of Easeus can answer this question since what the surface test does exactly is locked behind whatever information the UI displays. – Ramhound Jun 13 '22 at 19:38
  • It's a good question, and that would also make me suspicious of any results it gives. However as EaseUS Partition Master is proprietary, it seems to be difficult to find any independent information on its workings (and reliability of results). – MiG Jun 13 '22 at 20:02
  • And although it might be slower, considering these disks will be part of an array, wouldn't it be better to go for more reliable test results and for example use Western Digital's own **Data Life Guard Diagnostics**? This one isn't even brand specific, it can apply its tests to any drive. And yes, a full test would probably require about 32 hours per disk in your case, it might outrun waiting for a solid answer on EaseUS (especially if you run both drives' tests in parallel) :) – MiG Jun 13 '22 at 20:04
  • "*Data Lifeguard Diagnostic for Windows has been DEPRECATED and replaced with the Western Digital Dashboard utility. Data Lifeguard Diagnostic for Windows is no longer supported. Please use the Western Digital Dashboard utility*" --> Is it safe to use an old deprecated utility on modern hardware? Also, it took ~30-32 hours with EaseUS. If WD utility takes the same amount of time, it would indirectly clear EaseUS of suspicion for being suspiciously too fast. – InfiniteLoop Jun 13 '22 at 20:17
  • In that case I would use the WD Dashboard utility :) And I was under the impression your test was performed in a fraction of that time, in which case I would definitely be suspicious. I'm a bit surprised you would consider a 32 hour test to be too fast! – MiG Jun 13 '22 at 20:40
  • 1
    On 18 TB disk - yes. I mean, I showed the math behind my suspicion - if the drive writes AND reads to a sector as part of the test, and the speed of write is 200 MBs, it should perform the test at most at 6,000 MB/minute, and yet it was testing at 12,000 MB/minute as if it wasn't both writing AND reading, but doing ONLY ONE of those things (unless it can simultaneously do write and read BOTH at 200 MBs **simultaneously**, but AFAIK that's not how drives work). – InfiniteLoop Jun 13 '22 at 21:13
  • I'd stick to MB/s as the default currency, that's how performance is reported. You're right, even 200 MB/s for one of these (reading or writing), if sustained over long periods of time, is already high - having both read and write actions would be too much. Note that DLG (and I assume Dashboard) as their default extended test do 'long read', which means writing is not part of the test. Now I'm doubly curious what's going on with EaseUS. – MiG Jun 13 '22 at 22:41
  • So, I tried WD utilities with my 2nd drive - both the deprecated one and the new one. New dashboard couldn't even do a SMART scan (neither short nor extended) by failing with unspecified error within seconds. The deprecated app did start a scan but gave "too many bad sectors" error within like a few minutes, which I doubt is realistic. I wonder if they are having issues interfacing with the hard drive over the SATA-USB3.0 adapter. I don't have any means to plug natively into SATA (only have a laptop), so I guess I'll stick with EaseUS for now and maybe try Macrorit or smth else later. – InfiniteLoop Jun 13 '22 at 23:20
  • It is curious that both drives made some pretty loud sounds (clicking and screeching/revving up noises for like a minute or two when I plugged them in for the very first time. Quite operation after that (for 10s of hours), but the first minutes very loud. Is that normal? – InfiniteLoop Jun 13 '22 at 23:22
  • I know that DLG does a short drive scan preceding the long one, it could simply be reading what the drive itself reports as the number of bad sectors. Hence quickly rejecting the drive. – MiG Jun 14 '22 at 07:14
  • I would say no, clicking and screeching/reving noises sound like there's severe mechanical damage (you could record and share it so we can confirm, but my WDs don't do that). Did these drives get dropped? – MiG Jun 14 '22 at 07:16
  • 1
    SMART reports zero bad sectors in Crystal Info. I didn't drop the drives, but I don't know if they got dropped during shipment. BH Photo didn't package them well. THat's why I am checking these drives in the first place. – InfiniteLoop Jun 14 '22 at 19:42

0 Answers0