Hive-IO Storage

Hive-IO Storage

All Storage Options are included in the Hive-IO Software

For Light Workloads

RAM as Primary Storage (RaaPS)

storageRAM as Primary Storage (RaaPS) is built into the Hive-IO Stack. Deduplication and compression included as a standard feature. The RaaPS includes:

  • Baremetal access to RAM
  • RAM Disk seek time 1/1000th of a millisecond
  • Suited for low latency workloads
  • RAM Disk capacity looks 2-10 times larger
  • 5 million IOPs throughput availability
  • Driver-oriented, not protocol oriented
  • Significantly different from competitors, with storage acceleration before the wire

Local Attached Storage (Hyperconverged)

  • Handoff from bare-metal single disk or RAID
  • Deduplication, compression and cache for all storage configurations
  • Disk capacity presented 2-10 times physical size
  • Disk seek time is .1 ms • 2.5 million IOPs thruput
  • Driver-oriented, not protocol oriented
  • Significantly different from competitors, storage acceleration before the wire

Hyperconverged Remote NAS (Network Attached Storage)

  • NAS Client Connectors available in Hive-IO: NFS, CIFs, Ceph
  • Used for Multi-datacenter storage, High Availability

For Enterprise Workloads

Ceph File System (RBD)

  • Ceph Client plugs into Ceph Open Source Ecosystem
  • Can be configured to consumes S3 blocks, CEPH RBD
  • Can be configured to allow local Ceph nodes on HiveIO Fabric to discover other nodes in cluster
  • Creates a high available storage fabric ecosystem with deduplication, compression and caching (access time maximum .1 ms and requires ½ traditional disk space)
  • No latency (like with NAS) as everything looks like a local write
  • Perfect for large, complex workloads (HADOOP, Oracle, MS DB, Data Archive)
  • Partner with HiveIO Live migration to create most cost effective solution
  • Redundant ubiquitous storage – can grow as needed

Amazon Gateway Appliance

  • Mounted inside Hive-IO as a NFS
  • Read/ Write as S3 block and sent to S3 storage
  • Encrypted end-to-end to and from Amazon
  • If total failure occurs, Hive-IO offered a 1 h RTO resulting in no data loss
  • S3 encrypted per client
  • Using HiveIO read/write, compression, deduplication and redundancy across public and private internet
  • Dark storage everywhere
  • Full VDI and VSI workloads
  • Heavy Hadoop and Oracle workloads
  • Cost, throughput and availability are where we market and compete
Download the Hive-IO Reference Architecture - Receive 5 Test Licenses