
Modernize MongoDB Deployments
Leverage Pavilion Storage to Deliver a Rack-Scale Disaggregated Storage Service for MongoDB
- Independently scale compute and storage using a Pavilion array as centralized rack scale storage for MongoDB Clusters
- Improve performance by offloading storage processing to a dedicated appliance
- Enable rapid deployment of new secondary nodes, or full cluster copies for test/dev/analytics purposes
- Data Management services allow for new levels of operational efficiency
- Seamlessly integrate into the management environment by leveraging Pavilion REST APIs
- Reduce the amount of raw flash storage required to deploy MongoDB
Scale-Out MongoDB Cluster using Direct-Attached Storage
Key Benefits
- 120 GB/s Read, 60 GB/s Write
- 100 us Latency
- 20m 4K Random Read IOPS
Pavilion Overview
- Dual-Parity RAID
- Thin Provisioning
- Snapshots & Clones
Operational Benefits
- Reduce management complexity by moving storage to a shared appliance where the storage pool can be centrally managed.
- Deploy secondary nodes without copying data over the network by leveraging instant, zero-space writeable Clones
- Snapshots and Clones can be presented to any connected host and are not restricted to where the master volume is being used
- Isolate data backup and snapshot management and create zero-space instant backups of entire MongoDB clusters
- Manage the Pavilion Array through a UI interface, and/or supported by REST API, including an OpenStack plugin
Leverage Pavilion’s Intelligent Copy Management Capabilities to Simplify MongoDB Cluster Operations
Infrastructure Benefits
- Simplify the environment by leveraging a single, high-speed storage platform for MongoDB
- Deploy ‘storage-less’ 1U servers to deliver more container density per rack
- Save cost in several areas including hardware acquisition, rack space, power and cooling
- Pavilion’s Thin Provisioning support allows for less raw flash capacity to be installed
- Transition to Infrastructure 3.0
Dramatically Increase Compute Density per Rack for MongoDB Clusters