I HIGHLY recommend XenForo on SSDs!

[Jaxel]

My server resulted in 1.1 GB/s
http://www.knownhost.com/managed-ssd-vps-hosting.html

NOT dedicated... so I think thats pretty ******** fast.

 

[MattW]

My server resulted in 1.1 GB/s
http://www.knownhost.com/managed-ssd-vps-hosting.html

NOT dedicated... so I think thats pretty ******** fast.

This is from one of their SSD4 VPS's that I help manage

16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 1.25163 s, 858 MB/s

 

[Carlos]

I thought that SSD's are deemed as "unreliable?" Or has that improved since it was introduced?

 

[CTXMedia]

16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 5.29103 s, 203 MB/s

Single Intel 520 (with 30% over provisioning) driven by an AMD 6 Core CPU.

I thought that SSD's are deemed as "unreliable?" Or has that improved since it was introduced?

On the contrary, they fail more reliably - in that they stop being writeable; your data remains intact and accessible. Sure, the cheaper ones are prone to complete failure - just like HDDs - but the early days of really unreliable SSDs are pretty much gone and with an increase in MB/£ they are becoming more affordable at larger sizes too.

 

[Jaxel]

The thing about SSDs is each sector on the drive has a finite number of writes it can do to that block; usually around 100,000. Now this may not seem like much, but thats 100,000 for EACH block. If your drive has 10 million blocks, thats a lot of writing you can do. For thumb drives, with smaller capacities this was an issue; since they would reach end of life faster... in addition most solid state drives degrade in performance once 50% of the drive is consumed.

However, operating systems these days have a function called TRIM. While TRIM is normally used to "zero" out a sector from the drive and tell the OS that a section is writable again, it is also used to inform when a block is no longer writable... and then your OS automatically works around the dead sector.

 

[Luke F]

On the contrary, they fail more reliably - in that they stop being writeable; your data remains intact and accessible. Sure, the cheaper ones are prone to complete failure - just like HDDs - but the early days of really unreliable SSDs are pretty much gone and with an increase in MB/£ they are becoming more affordable at larger sizes too.

I can't seem to find it, but there is a forum thread somewhere where a bunch of SSDs were pushed to their write limit and beyond, and only a handful failed like that - the vast majority became completely inaccessible

 

[Jaxel]

I can't seem to find it, but there is a forum thread somewhere where a bunch of SSDs were pushed to their write limit and beyond, and only a handful failed like that - the vast majority became completely inaccessible

Dont buy OCZ.

 

[Luke F]

Dont buy OCZ.

Yeah OCZ (and anything with a sandforce controller) are trash, but this test included things like crucial m4's and samsung 830's and they met the same fate

It's not really a big issue though, because any good SSD will have fully predictable failure via SMART

 

[DRE]

My server resulted in 1.1 GB/s
http://www.knownhost.com/managed-ssd-vps-hosting.html

NOT dedicated... so I think thats pretty ******** fast.

I use knownhost managed ssd vps hosting. It's aight. Huge KnownHost thread here: http://xenforo.com/community/threads/knownhost-server-configuration.56303/