-2

Until recently, Dell et all would only provide slc drives for servers.

It appears the winds have shifted and now the cheaper mlc drives are avail for servers.

In a sql environment, will mlc give acceptable life? And in event of fail, well the raid mirror stay up long enough for us to swap out the dead drive?

Looking for real world experiences of this kind.

Thx!

Jonesome Reinstate Monica
  • 5,347
  • 9
  • 54
  • 80
  • 1
    Defiune "sql server". I know one that writes tons (around 400gb per week) - and another one that only adds some gb every week or so. That goes into "provide the information for an answer if you want an answer". – TomTom Dec 07 '14 at 17:11
  • possible duplicate of [Are SSD drives as reliable as mechanical drives (2013)?](http://serverfault.com/questions/507521/are-ssd-drives-as-reliable-as-mechanical-drives-2013) – ewwhite Dec 07 '14 at 17:21
  • No it doesn't. The only answer that matters is that no matter what you do on your server, you have a warranty and the vendor will replace a failed disk. – Basil Dec 07 '14 at 17:22

3 Answers3

2

Depends on what you're doing to the SSDs. Today, MLC versus SLC doesn't matter as much as buying the appropriate drives with the right attributes for your use case and environment. If that means "write heavy", "read optimized", "low latency", whatever...

See: Are SSD drives as reliable as mechanical drives (2013)?

But why complicate your thinking? If a drive fails, does it matter why? Use RAID, take and test backups and ensure you monitor your hardware and RAID array status. Or better yet, use drives from your server hardware vendor that match the characteristics of your workload.

Community
  • 1
ewwhite
  • 194,921
  • 91
  • 434
  • 799
2

Buy the SSD drive that matches your expected workload.

From the Dell SSD FAQ:

The use of an endurance management algorithm ensures that sufficient Program/Erase (P/E) cycles are available for the warranty time period of the drive. The firmware will limit writes if a drive is written heavily.

If you buy the a drive not intended for a write heavy IO load, but your usage pattern is, the firmware will limit writes to ensure you don't kill the drive while still under warranty... That will kill your application instead, but that's not Dell's problem.

HBruijn
  • 72,524
  • 21
  • 127
  • 192
  • Thank you. This is the most useful. Our app dies.... now I have to figure out if our writes per day might threaten our application. I don't suppose there is a utility to see the firmware limit coming before it hits? – Jonesome Reinstate Monica Dec 07 '14 at 21:15
  • No idea if and how that's actually implemented. One assumes it is à metric that can be queried using S.M.A.R.T the killing of applications was figurative , you simply at not get the performance you want and had before. – HBruijn Dec 07 '14 at 23:40
  • Upon reading that PDFs I noticed it is from 2011 so things may not be that bleak. – HBruijn Dec 07 '14 at 23:47
1

Assuming you keep it under warranty, if it ever fails, they'll replace it. The expected lifespan is their problem.

Basil
  • 8,811
  • 3
  • 37
  • 73