1 |
Am 08.07.2013 23:42, schrieb Alan McKinnon: |
2 |
|
3 |
> If it makes you feel better, then by all means go through the motions |
4 |
> . |
5 |
> |
6 |
> For my money, I reckon that's exactly what it is - motions and ritual. I |
7 |
> havew any anecdotal evidence to back it up, but it's fairly strong |
8 |
> anecdotal evidence: |
9 |
> |
10 |
> Over the last 5 years, the team I'm in, the teams we work closely with |
11 |
> and the Storage guys have commissioned >1000 pieces of hardware and |
12 |
> probably more than 4000 drives, the vast majority from Dell. I have no |
13 |
> idea what burn-in Dell applies, if any. We've had our fair share of |
14 |
> infant mortality failures, prob ably less than 20 in 5 years. And here's |
15 |
> the kicker - every single one failed in production. |
16 |
> |
17 |
> Most of that hardware, and ALL of the SANs, went through heavy |
18 |
> pre-deployment testing. Usually, this means cloning the -dev system onto |
19 |
> it and running the crap out of it for a decent length of time. Once the |
20 |
> techies were happy, install the production version and switch it on. |
21 |
> |
22 |
> I conclude that the likely reason we only found failure in prod is that |
23 |
> only prod gives a decent viable test that approximates real life and dev |
24 |
> is always a mere simulation. It's not usage that kills a few drives |
25 |
> early, it's the almost random pattern of disk access that you get in |
26 |
> real life. That tends to shake out the weak links better than any test. |
27 |
> |
28 |
> However, this is all anecdotal so use or discard as you see fit :-). I |
29 |
> no longer worry about data loss as we have 4 hour warranty turnaround |
30 |
> SLAs in place and company policy is to only deploy storage that is |
31 |
> guaranteed to survive loss of any one drive in an array. |
32 |
|
33 |
Thanks for that, good point :-) |
34 |
|
35 |
Stefan |