If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Display Modes |
#1
|
|||
|
|||
Some observations from Backblaze Q1 stats.
https://www.backblaze.com/blog/hard-...rates-q1-2017/
1. Seagate ST6000DX000 6TB "desktop" drives -- seem to be significantly more reliable than the WD 6TB WD60EFRX NAS drives 2. Seagate 8TB ST8000DM002 "desktop" drives seem to be somewhat more reliable than the Seagate 8TB ST8000NM0055 "Enterprise" drives. Perce |
#2
|
|||
|
|||
Some observations from Backblaze Q1 stats.
Percival P. Cassidy wrote:
https://www.backblaze.com/blog/hard-...rates-q1-2017/ 1. Seagate ST6000DX000 6TB "desktop" drives -- seem to be significantly more reliable than the WD 6TB WD60EFRX NAS drives 2. Seagate 8TB ST8000DM002 "desktop" drives seem to be somewhat more reliable than the Seagate 8TB ST8000NM0055 "Enterprise" drives. Perce You would have to normalize their chart to show drives having the same effective I/O load over the same duration to know which were the winners and losers. That chart by itself is useless. We're supposed to assume that "drive days" means each drive brand receives the same I/O load. We don't know and they don't say. How does Seagate's 267 failures in 34540 drives come to a 3.27% failure rate? I get 0.77% (267 / 34540 * 100). Presumably this is over more than a 1 year test duration so the annual failure rate would be even lower. Simple math on the worst Seagate shows a 8.8% failure rate (over how long is not divulged) instead of their cited 35%. For the annualized failure rate to be so much higher for Seagate means those drives received a lot higher I/O load. If Seagate is so bad, why does Backblaze still buy so many of them? The article has a link to supposedly the hard drive statistics data. Really? I went there and there was no statistical data, no spreadsheet, nothing there to do a separate and accurate statistical analysis. May the 2017 .zip files had that data but all I got was a constant waiting for their site to respond (deliver the file). Nowhere do I see mention of burn-in to eliminate bias due to infant mortality rate. Drive vendors, RAID vendors, and the industry talk about it but not Backblaze. Does Backblaze run the new drives through a burn-in period? Does whomever they buy the drives do a burn-in? |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
"Backblaze Hard Drive Stats for 2016" | Lynn McGuire[_3_] | Storage (alternative) | 4 | February 3rd 17 09:12 PM |
Backblaze pods harder on Seagate / Lawsuit | Ed Light | Storage (alternative) | 3 | May 26th 16 03:57 AM |
BackBlaze Storage Pod 5.0 - 180 TB | Lynn McGuire[_2_] | Storage (alternative) | 4 | December 25th 15 10:20 PM |
Backblaze dumps hard drive data | Neill Massello[_3_] | Storage (alternative) | 9 | February 20th 15 05:45 PM |
Backblaze on their experience with enterprise drives | miso | Storage (alternative) | 1 | December 8th 13 05:11 PM |