Jump to content
jbartlett

DiskSpeed, hard drive benchmarking (unRAID 6+), version 2.8

476 posts in this topic Last Reply

Recommended Posts

Vendor:  Western Digital
Model: WDS100T3X0C-00SJG0  -- Western Digital Black SN750 1TB NVMe
Serial Number: 201802800032
 

Revision: 111110WD

Share this post


Link to post
2 hours ago, jms2321 said:

Vendor:  Western Digital
Model: WDS100T3X0C-00SJG0  -- Western Digital Black SN750 1TB NVMe
 

Revision: 111110WD

Hrm. Nothing stands out as being potentially invalid

Share this post


Link to post
Posted (edited)

Thank you @jbartlett

 

This plugin looks amazing.

 

I uploaded a picture of, and info relating to, my 1TB Intel 660p NVME SSD to the database.

 

For some reason, my 16GB thumb drive stick shows as a 15GB thumb drive:

100 Series/C230 Series Chipset Family USB 3.0 xHCI Controller
Super Micro Computer Inc (Intel Corporation)
USB controller

Drive ID: sda (flash)
Vendor: SanDisk'  
Model: Cruzer Fit
Serial Number: XXXXXXXXXXXXXXXXXXXX

Revision: 0
Capacity: 15GB
Logical/Physical Sector Size: 512/512

Any idea why ? Not that this is an issue, I am just curious. 

 

Best,

OP

Edited by Opawesome

Share this post


Link to post
17 hours ago, Opawesome said:

For some reason, my 16GB thumb drive stick shows as a 15GB thumb drive:

Any idea why ? Not that this is an issue, I am just curious. 

Open a telnet prompt and type in "lshw -c disk" which will list most of the storage media attached. Look for your USB media. Mine looks like the following

  *-disk
       description: SCSI Disk
       product: USB Flash Drive
       vendor: Lexar
       physical id: 0.0.0
       bus info: scsi@0:0.0.0
       logical name: /dev/sda
       version: 1100
       serial: AA00000000000489
       size: 29GiB (32GB)
       capabilities: removable
       configuration: ansiversion=6 logicalsectorsize=512 sectorsize=512
     *-medium
          physical id: 0
          logical name: /dev/sda
          size: 29GiB (32GB)
          capabilities: partitioned partitioned:dos
          configuration: signature=c3072e18

Look at the size on both the disk & medium. If they match, your drive was mislabeled. If the Medium is smaller, than the entire device isn't partitioned to it.

Share this post


Link to post
Posted (edited)

getting the following error in the docker log.

 

"ERROR","http-apr-8888-exec-1","06/25/2020","08:45:34","",";Lucee was not able to load a Agent dynamically! You need to load one manually by adding the following to your JVM arguments [-javaagent:""/opt/lucee/server/lucee-server/context/lucee-external-agent.jar""];lucee.runtime.exp.ApplicationException: Lucee was not able to load a Agent dynamically! You need to load one manually by adding the following to your JVM arguments [-javaagent:""/opt/lucee/server/lucee-server/context/lucee-external-agent.jar""]

 

Diagnostic file emailed to you!

Edited by jms2321
added detail

Share this post


Link to post
2 hours ago, jms2321 said:

getting the following error in the docker log.

 

"ERROR","http-apr-8888-exec-1","06/25/2020","08:45:34","",";Lucee was not able to load a Agent dynamically! You need to load one manually by adding the following to your JVM arguments [-javaagent:""/opt/lucee/server/lucee-server/context/lucee-external-agent.jar""];lucee.runtime.exp.ApplicationException: Lucee was not able to load a Agent dynamically! You need to load one manually by adding the following to your JVM arguments [-javaagent:""/opt/lucee/server/lucee-server/context/lucee-external-agent.jar""]

 

Diagnostic file emailed to you!

I see that in mine too, I'll investigate. I'm assuming the app works fine otherwise?

Share this post


Link to post

Honestly, I died when it started benchmarking the drives on the second hba.  I will rerun it tomorrow when I have time to monitor the system closer.

 

Share this post


Link to post
On 6/5/2020 at 8:54 PM, jbartlett said:

I'll add some debugging to the scanning process that'll hopefully help pinpoint your issue.

 

Did this debugging get added?  I'm still not having drive images download automatically.

Share this post


Link to post
On 7/7/2020 at 8:11 AM, Alexstrasza said:

 

Did this debugging get added?  I'm still not having drive images download automatically.

Not yet. I haven't had time to do so yet. Personal time on the PC has been limited of late. I'll post an update when it is.

Share this post


Link to post
On 7/7/2020 at 8:11 AM, Alexstrasza said:

 

Did this debugging get added?  I'm still not having drive images download automatically.

I've made some changes to the image search routine to make a log. If you still do not get any images, please comment with the time that you ran your scan (with time zone) so I can find your specific logs.

Share this post


Link to post
Posted (edited)

Docker Version 2.8 has been pushed

  • Don't include floppy devices
  • Multiply the SMART LBA Read/Written values by the Logical Sector Size instead of the Sector Size
  • Add more validation logic to clean up Vendor/Model combinations

I have rewritten from the ground up the Hard Drive Database website at http://strangejourney.net/hddb/ to display meaningful benchmark graphs. When viewing benchmark data for a particular drive, a graph line is drawn for each unique drive's most recent benchmark is displayed. If 100 people own a given drive and they've all done one or more benchmarks against that drive and uploaded them, the HDDB would have 100 graph lines for those drives. It gives you an idea as to how stable that model is on performance.

Take Western Digital model WD30EFRX. It has two revisions, 80.00A80 & 82.00A82. The 1st one is has a slightly wide speed variance up to 30MB/sec at the start of the drive but the 2nd revision has two distinct groupings which indicates that something changed between the two sets but no indication of which drive you got.

 

image.thumb.png.85a083cb9f2f470a34f1de2cb9db954e.png

 

This graph for a WD40EFRX is indicative of frequent drive failures as well as a fairly wide variance of possible speeds (and also looks like multiple groupings).

image.png.6967c4dbb5764ebff484ddf11c3c3134.png

 

The speed grouping could be the result of the storage controller it was attached to but I kinda doubt that.

Edited by jbartlett

Share this post


Link to post

I just noticed that the two graphs with the multiple groupings both have the same revision number which tells me that Western Digital used the same circuitry with different number of platters. A quick query shows this same trend over multiple vendors. There might be some insight into seeing how the same revision fairs on different drive media configurations.

Share this post


Link to post

Wow, this is turning into quite the valuable resource!

 

If you can, maybe the graphs could have a view option to feel less busy by widening the normal ranges into a light background, with the current upload as a single color line. The end result would be a black and white greyscale graph, with a colored line either passing through the normal white range or deviating into the grey or black regions.

 

The graph would start as pure black, each additional data point would add a brightness level divided by the total number of lines plotted. So if all the curves happened to pass through the same point, it would be pure white, and if none of the curves hit that point, it would be black. The first curve would be pure white, 2 curves would be 50% unless they intersected into pure white, et. cetera. If you only have a handful of curves it probably wouldn't be a very useful, but when you have 100's of data points it should start to resemble something intuitively useful.

 

Share this post


Link to post

i opened a Brand Image database, then when i click on particular disk, it throws 404 error.

image.png.9a185c48e2a9e5bf402811a7c68a8b31.png

Share this post


Link to post
14 hours ago, uldise said:

i opened a Brand Image database, then when i click on particular disk, it throws 404 error.

Sorry about this. This has been resolved.

Share this post


Link to post
On 7/31/2020 at 7:18 PM, jonathanm said:

Wow, this is turning into quite the valuable resource!

 

If you can, maybe the graphs could have a view option to feel less busy by widening the normal ranges into a light background, with the current upload as a single color line. The end result would be a black and white greyscale graph, with a colored line either passing through the normal white range or deviating into the grey or black regions.

 

The graph would start as pure black, each additional data point would add a brightness level divided by the total number of lines plotted. So if all the curves happened to pass through the same point, it would be pure white, and if none of the curves hit that point, it would be black. The first curve would be pure white, 2 curves would be 50% unless they intersected into pure white, et. cetera. If you only have a handful of curves it probably wouldn't be a very useful, but when you have 100's of data points it should start to resemble something intuitively useful.

 

That's basically describing a heatmap. Interesting concept that I'll likely explore - notably because I added a graph with all of the benchmarks using the same revision & RPM and it was just a hot mess of lines.

Share this post


Link to post
4 minutes ago, jbartlett said:

just a hot mess of lines

Yeah, that's what I anticipated based on what you posted, the mess of lines confuses rather than informs. Maybe even explore what a 3D plot would look like, with the current result draped across the 3d landscape, that way you could easily show abnormalities with green where the current plot follows the trend, blending through yellow and then red the farther away from normal you get.

Share this post


Link to post
16 hours ago, jonathanm said:

Yeah, that's what I anticipated based on what you posted, the mess of lines confuses rather than informs. Maybe even explore what a 3D plot would look like, with the current result draped across the 3d landscape, that way you could easily show abnormalities with green where the current plot follows the trend, blending through yellow and then red the farther away from normal you get.

I have two data points, drive location and speed. I'd need a 3rd data point for a 3D representation, what would that be?

Share this post


Link to post
2 hours ago, jbartlett said:

I have two data points, drive location and speed. I'd need a 3rd data point for a 3D representation, what would that be?

Frequency of appearance in the database of that location and speed.

Share this post


Link to post
3 hours ago, jonathanm said:

Frequency of appearance in the database of that location and speed.

That would be better represented by a color change. The heat map will be a something like an 800x200 grid where the drive size and read speeds are scaled to fit in that. The more times a spot gets hit, the brighter that spot would become. The line between the 11 spots (scanned every 10% plus the start) would be interpolated. Here is a demo that shows a heat map tracking temperatures throughout the year.

Share this post


Link to post

Heatmap data has been added to the Hard Drive Database and is displayed for any drive model that has more than 25 drives benchmarked & uploaded.

 

The 9th most popular drive based on uploaded benchmarks is the 8TB Seagate ST8000VN0022. Looking at the graph shows and odd dip at the 4.8GB mark but there are enough seemingly normal scans that you'd think that it might be a fluke but the heatmap scan shows the majority of the drives do have a dip at this point.

 

http://strangejourney.net/hddb/

 

image.thumb.png.cbf4b84f6dd878a17a70c36e24e6a4ad.png

Share this post


Link to post

Hi

 

I just built my first Unraid box and installed this DiskSpeed Docker container to test out the 3 new drives i purchased and installed.

 

I did my first test of each drive and the results are below. From other images ive seen in this forum, the parity drive looks to be about right in regards to speed and the slow drop off in speed. Disk1/DIsk2 however look quite different and im just wondering if that's normal?

 

The 3 drives are all WD Red 4TB (WD40EFAX)

The parity drive is plugged into a SATA 6.0Gb/s port

Disk1/Disk 2 are plugged into SATA 3.0Gb/s ports

The motherboard is an Asus H81M-E

 

Also when i complete the test on Disk1/Disk2 the following note is displayed "Bandwidth was capped on drive Disk 1 (sdd)"

 

So does it look like there is an issue with the drives or am i doing something wrong?

 

Thanks for any help.

 

benchmark-speeds.png

Share this post


Link to post
2 hours ago, Jmulligan said:

The 3 drives are all WD Red 4TB (WD40EFAX)

Those are SMR, they will only show the correct performance after being written completely at least once.

Share this post


Link to post

If you notice how the drives have a flat line? That's what the program looks for to determining if bandwidth was capped - or if it seems the drive can output data faster than the drive controller or connection can handle.

 

As per tee-tee jorge's comment, run a pre-clear on those drives and you should get graphs that match the trend on the Hard Drive Database

Share this post


Link to post

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.