Hacker News Reader

If you look, that drive model is also the most highly used by far. I think it's just a matter of the larger sample size / use time.

View on HN

Surely it doesn’t matter when you have 10,000s of drives? Aren’t you already at a large enough sample size? If it isn’t, what is the point of them publishing this every year? I don’t know the math of the matter though.

Yea I don't know. I'm not big on statistics either. I just noticed that the drives that did the worst were the ones that had the most usage overall.

That would probably be price ~ failure rate correlation.

> Surely it doesn’t matter when you have 10,000s of drives? Aren’t you already at a large enough sample size? If it isn’t, what is the point of them publishing this every year? I don’t know the math of the matter though.

I think drive age matters? I'm not clear if they cycle drives out at a certain age or just run them until they fail.

Also, if a drive is low enough in cost, then the additional cost of replacing an incremental 1% may be lower than the cost of acquisition of a more reliable drive.

Looks like the Segate 0007 are 1y old on average, where the 0008 are 44 days old on average.

The 12TB HGST are 220 days old on average. The Segate 12TB failure rates seem high, quite unfortunate as I own 6 of them.

Reply on news.ycombinator.com