The VNX series enables you to leverage a single platform for file and block data services. Centralized management makes administration simple. Data efficiency services reduce your capacity requirements. EMC annouce the VNX-F
Lets look at the VNX-F, what it is and what’s new.
VNX-F is an all Flash, block only offering with advantageous pricing compared with standard VNX and all flash pools. It is designed to fit into the EMC family of Flash offerings that includes XIO in the enterprise space and VNX-F in the mid-range. VNX-F offers a low cost all-Flash configuration with outstanding and proven availability.
So, lets look at everything that is new with the VNX-F.
- A VNX-F7000 with some specific capacity points based on 400GB SSDs. Well, EMC will change the capacity points and ADD 800 GB SSDs
- EMC added a new entry level VNX-F, namely the VNX-F5000. The VNX-F5000 will have a lower entry price, it will also have varying capacity points as well as supporting the 800 GB SSDs.
- Both the VNX-F5000 and VNX-F7000 have configurations at under $5/GB
- EMC will introduce the new 120 drive 3 RackU disk array enclosure (DAE) (aka drive shelf) and 1.6 TB SSDs. This allows the VNX-F to promote flash density – up to 46 TB usable /RackU.
A fully populated 120 drive 3 RackU DAE (code name Viking), with 1.6 TB SSDs, a RAID 5 (8+1) and the required hot spars results in 45.9 TB usable per RackU. However, the largest VNX-F configuration of 171 TBs usable requires a 25 drive Disk Processor Enclosure (DPE) + a 120 drive DAE. This results in a density of 28.6 TB usable/RackU.
Let’s get into this in more detail. Being an all flash configuration, you would expect good performance and low latency. Well the VNX-F delivers up to 400K at sub millisecond latency. The VNX-F will only support 400 GB, 800 GB and 1.6 TB SSDs in a 25 drive DPE, 25 drive DAE or 120 drive DAE. No other drive types are supports…and NO you cannot add other drive types later! ONLY block connectivity will be supported, i.e. FC, FCoE and iSCSI, NO file….and no, you cannot add file later!
The VNX-F is targeted at those customers who do not require typical VNX Data services. Of cause you have management suite and the app integrations. Both the VNX-F5000 and VNX-F7000 all flash configurations should be positioned in environments where price and density matter most and data services such as thin provisioning, deduplication, and compression are not requirements. Targeted applications include medium-sized databases, virtual application environments, and as all flash tier for SAN virtualization
So the VNX-F moves into the arena of Violin Memory, Pure Storage, Netapp and IBM
You may be asking, why is EMC introducing a VNX-F when they already have the XtremIO. According to EMC those two products are very different and focused at different markets.
XtremeIO is focused on high predictable performance with sub-millisecond response time even with data services turned on. VNX-F is a purpose built VNX all flash configuration that delivers high density, high performance (up to 400k IOPs) with configurations under $5 per usable GB to those customers who do not require advanced data services (like thin provisioning, deduplication, compression, etc.)
XtremeIO scales out whereas VNX-F scales up. What this means is that although XtremeIO scales to very large capacities (simply adding nodes to scale capacity and performance), they don’t come as far down as what VNX-F can provide. Meaning VNX-F has a smaller starting configuration, or entry price, but cannot scale as high as XtremeIO can.
XtremeIO offers those rich data services, like in-line deduplication, compression, writeable snaps, etc. without any impact on performance. Again, XtremeIO is about predictable performance even with services turned on. VNX-F has some optional services like D@RE, snapshots and remote replication, but it does not offer all the traditional rich VNX services like tiering, dedupe, compression, etc.