FlashMAX Connect Software
Many manufacturers overlook the big picture and focus solely on hardware. There are customers, such as the burgeoning hyperscale market, that create and manage their own environments to provide their own customized functionality. Unfortunately, this is a rather small subset of the market, and the majority of administrators are left wanting for traditional capabilities from their server-side flash.
One of the key differentiators of the Virident product line is their enhanced software application suites. Many users require SAN-like functionality, such as replication, high availability, failover, and sharing, to service clustered configurations and meet SLA's. Pairing these features with explosive flash performance has become a challenge.
The FlashMAX Connect software virtualizes memory and storage resources. This allows spanning flash capacity across multiple applications and servers by creating virtual flash storage networks with RDMA to provide the high availability, sharing, and replication features. This provides more flexibility and scalability, and increases efficiency by allowing storage to scale independently of compute resources. Users can integrate .5 TB to 24 TB of flash per server, and scale out by adding servers on an as-needed basis. The architecture has a side benefit of avoiding vendor lock-in.
FlashMAX Connect transforms normal off-the-shelf servers into distributed storage via three integral components. These components are deployed separately, or as an entire suite.
vHA - High Availability Flash Storage
vHA allows the creation of fail-over clusters by enabling synchronous replication among servers for the data stored on remote FlashMAX II SSDs. High-throughput and low-latency are key to retaining the performance benefits of flash, and replication is delivered by RDMA (Remote Direct Memory Access) to minimize latency (with compatible hardware, notably IB QDR or FDR with Mellanox ConnectX-2, X-3, or equivalent). Read traffic is serviced from the in-chassis SSD to minimize network overhead.
vHA supports failover clustering for Oracle, MySQL, PostreSQl, and other applications, and provides replication bandwidth up to 7GB/s per server and 48 microsecond latency for 4k transfers.
vShare - Shared Flash Storage
One of the most common complaints is that PCIe SSDs can create underutilized and isolated pools of storage. In many cases, the speed of one FlashMAX II device will outstrip the needs of a single server, and massive capacities up to 4.8 TB are just begging to be shared. Any underutilized resource represents wasted opportunity and cost, especially with the premium price structure of flash.
vShare extends block-level shared access of FlashMAX devices to multiple servers. This provides access to flash performance for servers that do not have in-chassis flash, and allows creation of high-performance active-active clusters with application-driven data mirroring, such as redundancy modes in Oracle ASM. The vShare driver also features multi-path capability to combat network-based failures.
RDMA once again provides the high-performance data transfer pipe to keep latency low and throughput high, and speeds reach 10 GB/s read + 10 GB/s write. The added 4k read/write latency is 20/40 microseconds, and up to 1,000,000 IOPS can be shared between servers. This is certainly helpful for those wishing to extend the benefits of flash outside of a single server.
vCache - SAN and DAS Caching
vCache configures the FlashMAX II as a fast front-end for SAN and DAS storage by caching hot data on the flash. vCache maximizes flash endurance by eliminating double mapping of block addresses and has three caching modes (write-back, write-through, and write around) to support varied applications and environments. Users can configure multiple cache pools on a single SSD to cache multiple LUNs. vCache can provide targeted caching by blocking certain processes, such as backup operations, from caching. This increases endurance and maximizes cache capacity.
The most impressive aspect is the combination of vCache with vShare, which provides access to cache pools located on other servers. This optimizes cache capacity in clusters by eliminating replicated cache data while providing easy acceleration to clustered applications. vCache provides 4k read latency at 85 microseconds, and write back cache latency weighs in at 16 microseconds.
Readers should research compatibility for each component to ensure interoperability with their environment.
PRICING: You can find products similar to this one for sale below.
United States: Find other tech and computer products like this over at Amazon's website.
United Kingdom: Find other tech and computer products like this over at Amazon UK's website.
Canada: Find other tech and computer products like this over at Amazon Canada's website.
Recommended for You
- We at TweakTown openly invite the companies who provide us with review samples / who are mentioned or discussed to express their opinion of our content. If any company representative wishes to respond, we will publish the response here.
Latest News Posts
- ISS crew arrives at Space Station after two-day trip
- Major cyber attack knocks out popular websites
- Cassini's final year of its epic voyage to Saturn
- Nintendo won't reveal full Switch specs until 2017
- Nintendo Switch may have extra power while docked
- BSI3H-6100 long time to post
- Patriot Memory LX 200GB microSD Memory Card Review
- ASRock problem SB fan1
- GA-Z170X-Gaming 7 - F20b BIOS BUG?
- Unable To Update AMD GPU Driver
- ENERMAX releases Steelwing aluminum case
- ENERMAX Platimax D.F. PSU is available now
- ENERMAX launches Revolution Duo power supply with DUOFlowTM design for active ventilation
- Razer unveils new Razer Blade Pro gaming notebook
- Nintendo Switch world premiere demonstrates new entertainment experiences from a home gaming system