Brand new drives, no data. They are connected to a SAS2LP controller and Norco 4224 hotswap bays. Correct me if i'm wrong but I should be seeing about 200MB/s initially based on other results i've seen. What do you guys think?

 

EDIT: If I click the eye, it shows a more reasonable 150-170MB/s for each drive, not sure why the WebUI claims slower.

Update: They are still going 150-160MB/s at 27% (4 1/2 hours at this point), barely any speed loss. I guess I was expecting their speeds to reduce much quicker than they are. Seeing most people in the Seagate thread getting ~20hr pre-reads so this seems about right.

From a read point of view they behave just like conventional drives and have a very respectable performance. From a write point of view, a single thread of sequential writes will be directed straight to the shingled bands, so expect good performance during pre-clears. The thing that hammers them is sustained random writes that fill up the persistent cache. If they seem to take a long time to pre-clear, it's simply because they are so big! I think the discrepancy you're seeing is because one reading is an instantaneous value and the other is a running average.

8TB Seagates performance curve looks like this:

Of course, the disk speed test only reads one disk at a time. Pre-clearing five simultaneously is asking a bit more of the system. The most I've ever attempted is four.

My guess is this has something to do with you running 6 at a time! It perfectly explains the lack of speed loss as the controller was the bottleneck up until this point.

8TB Seagates performance curve looks like this:

 

You must have a really good sample. 

I'm looking at this post (http://lime-technology.com/forum/index.php?topic=39526.msg386233#msg386233), and for some reason his pre-read is slower than write/post-read. Perhaps the pre-read just doesn't get that 200MB/s starting point. Maybe the speed just has to do with the preclear script itself.

 

 

His results:

---------------------------------------------------

Preclear

 

(Cycle 1, before power interruption)

 

- Initial Pre Read speed at the commencement of the preclear was ~170MB/s for ALL 3 drives.

 

- All three completed their Pre Read with a FINAL speed at 100% of ~80MB/s taking ~20 Hours.

 

- Initial zeroing speed of cycle 1 of 3 was ~200MB/s for ALL 3 drives.

 

- All three completed their Zeroing with a FINAL speed at 100% of ~138 MB/s taking ~16 Hours.

 

- Total Time for cycle 1 at this point was ~36 Hours.

 

- Initial Post Read speed was ~200MB/s for ALL 3 drives.

 

- All three completed their Post Read with a FINAL speed at 100% of ~115 MB/s taking ~16 Hours.

 

---------------------------------------------------

 

 

Of course, the disk speed test only reads one disk at a time. Pre-clearing five simultaneously is asking a bit more of the system. The most I've ever attempted is four.

 

It's the same speed whether I preclear 1 or 5, but it's not the 130-135MB/s the webUI is claiming. It's was ping ponging between 150-170MB/s when I went into the details, and right now at 35% it is ping ponging between 145-155MB/s.

 

Doesn't seem that far off from testdasi's results, so I think it's probably good enough.

 

My guess is this has something to do with you running 6 at a time! It perfectly explains the lack of speed loss as the controller was the bottleneck up until this point.

 

The first screenshot was with only 1 going, the second was with all 5 going. I wanted to show the speed was the same regardless of how many preclears I was doing. Sorry for the confusion!