Would you use consumer SSDs in a server?

[ShintaiDK]

Specifically, would you use any of the following:

Samsung 840 PRO
Samsung 850 PRO
Crucial MX100

I lean heavily towards the MX100s because they are cheap and for about the same price I can get one size bigger, which would help with the writes and endurance. Plus, they have capacitors, similar to the enterprise SSDs.

Also, would you pair different drives (models/manufacturer) but same size in RAID1 arrays (C224 chipset)? I'm thinking that it might save me in case of bugs one might experience in 24/7 use in consumer drive firmware.

The server would run MSSQL. The writes would be 2-5 GB per day.

Absolutely. We have been using consumer SSDs for quite some years now in servers. Back then it was easier due to the lack of enterprise class SSDs. But we still do it today and the failure rate is pretty much nonexistant.

 

[Chicken76]

Absolutely. We have been using consumer SSDs for quite some years now in servers. Back then it was easier due to the lack of enterprise class SSDs. But we still do it today and the failure rate is pretty much nonexistant.

What brands/models do you use? (or would use if you were to build new servers now)

 

[beginner99]

FYI the MX100 does not have power-loss protection. That is marketing BS. See the review of the M600:

http://anandtech.com/show/8528/micron-m600-128gb-256gb-1tb-ssd-review-nda-placeholder

The client-level implementation only guarantees that data-at-rest is protected, meaning that any in-flight data will be lost, including the user data in the DRAM buffer. In other words the M500, M550 and MX100 do not have power-loss protection -- what they have is circuitry that protects against corruption of existing data in the case of a power-loss.

Also about power-loss:

http://lkcl.net/reports/ssd_analysis.html

Conclusion:

As was already said and you confirmed: An UPS is mandatory especially with consumer-grade SSDs.

 

[Chicken76]

Thanks beginner99. That analysis was a good read.

 

[PliotronX]

Absolutely. We have been using consumer SSDs for quite some years now in servers. Back then it was easier due to the lack of enterprise class SSDs. But we still do it today and the failure rate is pretty much nonexistant.

Lucky! I can't convince anyone to use them in their servers "oh I heard you need expensive kinds of SSD's." The SSD makers are losing out on a potentially large market.

 

[beginner99]

Thanks beginner99. That analysis was a good read.

FYI I still think a consumer SSD is a good fit for the right server application and data. i would not do it with highly important sensitive stuff (like your core business) but for some not so crucial stuff makes sense.

IMHO it makes also sense for databases contrary to other opinions in this thread. in fact it makes the most sense there due to many small random read and writes in which ssd shine and hdds are crap.

As said have an UPS and regular back-ups and before deploying actual test that you back-up procedure works, eg. do a restore at least once and document the procedure to every single detail.

 

[ShintaiDK]

What brands/models do you use? (or would use if you were to build new servers now)

We used Intel until they started with the Sandforce crap. Then we used Crucial since.

 

[ShintaiDK]

Lucky! I can't convince anyone to use them in their servers "oh I heard you need expensive kinds of SSD's." The SSD makers are losing out on a potentially large market.

The SQL people didnt take more than a few seconds to convience

 

[Emulex]

I use samsung 830/840 pro's under esxi 5.1 with lsi megaraid controllers (raid-1) with 30% OP and it is 1000% solid since the day they were installed to present. no latency issues!