Jump to content

Recommended Posts

Posted

As per title has anyone had a play with the Lenovo SA120 disk array?

 

ThinkServer SA120 Direct Attached Storage |Direct Attached Storage SolutionsÂ| Lenovo US

 

It looks pretty good and very cheap for a lab setup at £500 for the dual controller/PSU model.

We are planning to test a setup with 2012R2 storage spaces and this as a JBOD

 

BT Business Direct - Lenovo ThinkServer SA120 Storage Enclosure (70F10002UK)

 

Or maybe has anyone else played with a different cheap array/JBOD?

  • Thanks 1
Posted (edited)

I'll be looking at Jbods in the next month and that is _really_ well priced!

 

Thanks for the find!!!!

 

* surely the price has got to be wrong? The single PSU single controller version is $2800 on the Lenovo website!

Edited by PotNoodleTech
Posted (edited)

Quick google on the single IOCC single PSU version:

 

 

70F10003UK Lenovo ThinkServer SA120

261.66

 

 

Lenovo ThinkServer SA120

257.20

 

 

Lenovo ThinkServer SA120

£260

 

 

Ebuyer:

Lenovo ThinkServer SA120 70F1 Storage enclosure 12 Bay 2U Rack Server - Ebuyer

293.74

 

We are going to test the single controller version, and if it works to plan we will buy another one and use the controller and PSU from that..

We only lose about £20-30 doing it that way and get a spare enclosure!

Edited by burgemaster
Posted (edited)

Wow, that's pretty darn cheap - especially as it is certified for Windows 2012/2012 R2 for use with their Storage Spaces/Failover Clustering!

 

Lenovo ThinkServer SA120 JBOD Review | StorageReview.com - Storage Reviews

 

Good review there.

 

This can't be used to create a high availability cluster with two hosts though as each controller only has one port (and one daisy chain to more storage).

 

It would make the ideal "lab" set up though!

 

Can it not be used with a single SAS channel to each server? Microsoft have certified it to work with the technology...

Edited by localzuk
  • Thanks 1
Posted

It may be me being a newbie as I'm learning all about this HyperV and in particular Jbods but I thought to have two inputs per controller (not output/daisy chain) so you could plug two servers in to each controller and then if a controller fails, both servers will still be able to communicate with the other controller?

 

With that set up you have one server plugged into one controller, so if the controller fails the server loses total communication with the JPOD?

 

Though saying that I am looking carefully into whether dual controllers are actually worth the money as, frankly there's no moving parts on a controller so what are the chances it would fail?

 

It's all a steep learning curve!!

Posted
You would just add a second controller in the JBOD, to provide for this. Remember that the JBOD would be linked to a 'header' server - 2 SAS HBA cards could be used to connect to each controller on the JBOD, accounting for fault tolerance. Alternatively, connect two header servers to the JBOD, with each SAS card on each server connecting to a dedicated controller on the JBOD. These file servers could then be setup as a SoFS cluster, which would provide fault tolerance for any Hyper-V hosts connecting through to the file server cluster - if one file server dies, there is still another that is connected to the JBOD.
  • Thanks 1
Posted

Thanke MRJDH this is where I am a bit unsure. IT would be great if you could clarify a bit more.

 

From what I can tell, SoFS overcomplicates / is above what I need? All I need to do is provide shared storage to two HyperV hosts to give High Availability in HyperV - so Each HyperV host connecting directly to the JBOD via a single SAS card would provide this wouldn't it? I.e if one host dies, the other copes, if one sas interface dies it would take out the host that is connected to it but the other host would then cope. The biggest risk of failure would be total loss of the JBOD device in which case it would be time for VEEAM to flex it's muscles?

Posted
SoFS would over-complicate the setup you desire. A SoFS cluster would protect against one of the file servers connected to the JBOD going bang. With the other node of your SoFS cluster, at least one file server would still be connected to the JBOD.
Posted

I don't think there would be anything stopping you, but I can't imagine it's a setup that most would employ.

 

The single point of failure is still the JBOD, having 2 file server nodes in a SoFS cluster still won't help that, but having two 'header' servers connected to the storage at least covers you for the complete failure of one of the nodes connected to the JBOD.

 

Check out Aidan Finn's blog - Hyper-V don!

http://www.aidanfinn.com/

  • Thanks 1
Posted

The SA120 arrived today. Bit of a let down, it is full of dummy drive trays and doesn't come with any to install!

Cant find a part number or anything similar to buy!

 

Will have to call Lenovo

Posted
After Buying IBM/Lenovo servers for many years... None of them come with trays, The trays come attached to the relevant drives you buy. I believe we got some trays separately once before from a spares supplier.
Posted
After Buying IBM/Lenovo servers for many years... None of them come with trays, The trays come attached to the relevant drives you buy. I believe we got some trays separately once before from a spares supplier.

 

HP do the same thing, I think this is pretty typical for all manufacturers.

  • 4 weeks later...

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...