Jump to content
SSD

Warning purchasing USB hard drives

56 posts in this topic Last Reply

Recommended Posts

There was an interesting article by Mr. Antman (England) checking a Seagate Archive drive in a Drobo which failed d/t the SMS type file structure being used.  In summary the Write Head is larger than the Read Head, thus multiple writes to the HD means the data must be read (d/t size) then re-written.  Not sure how that might affect Parity ????

 

I have not seen that article, but it is not so much that the head is larger, the required track on media is larger. The track needed to write is wider than needed to read, so smr (shingled) overlap the writes, removing some the "extra" area written, since it is not needed to read. This relationship between write and read requirements is unlikely to change in future generations.

 

This article covers in depth many of the characteristics of the SMR drives.

https://www.usenix.org/system/files/conference/fast15/fast15-paper-aghayev.pdf

 

In summary, sustained random writes will not perform well. Random writes up within the capacity of the persistent cache, followed by lengthy (hours) idle time (no reads or writes) will achieve good performance.

 

Share this post


Link to post

Remember these are called "shingled" drives => the name is derived from the fact that the writes overlap each other, as c3 noted above -- think of the shingles on a roof (this is why they named the technique "shingled").    The drives perform very well for reads (no different than standard PMR drives); but to write anything on a specific track requires re-writing the entire track as well as every track after it up until the next "buffer" track (where there's no shingling).    This could result in TERRIBLE write performance, except there is an area reserved for "buffering" all of these random writes, known as the "persistent cache", which buffers writes to a non-shingled area, and then moves them to the shingled sections later (during idle time).  Seagate has done a very good job of mitigating these issues, as I've outlined in this thread: http://lime-technology.com/forum/index.php?topic=39526.0

 

Share this post


Link to post

Hi garycase/C3,

 

Many thanks for the responses and the detailed information.  Gary I read through your analysis per the post and tried to read danioj analysis as well (a bit over my head), but bottom line, I'll go ahead an preclear one of the disks at a time and see how they stack up. 

 

I really appreciate your responses to my questions.  Is there any advantage to taking the drive out of the case vs using USB3 to preclear?

 

Regards and Kudos!

 

Dave

Share this post


Link to post

I haven't done any USB-based pre-clearing; but my understanding is the newest version supports USB v3, so as long as your computer has USB3 ports, pre-clearing via USB should be just about as fast as a direct SATA connection -- and has the advantage that you'll know the drive is good before opening the case to remove it (thus voiding the warranty).

 

Share this post


Link to post

Hi Gary,

 

It's currently in the Pre-read phase at ~ 2 hours @ an average of 198 mb/sec so I guess it  works! ;D

 

Dave

 

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.