DSS Automation
Automation is mandatory to reduce the management cost of large storage systems. DSS allows your storage admins to manage over 10x more capacity as the day to day storage management and operational tasks are automated.
Adding storage capacity
DSS is scalable by simply adding storage nodes. Whenever storage capacity is added, the DSS system will automatically start consuming the new capacity. Unlike traditional systems where volume management is required to reconfigure RAID groups or volumes, DSS will start consuming the new capacity without having to reconfigure the access interface to the storage system.
Automated failure resolution
DSS is configured to survive multiple node or disk failures. Whenever a storage node would fail, DSS will automatically reconfigure the data spread to use available disks for new incoming data. Thus a subsequent write is immediately compliant again with the availability policy.
In the event of a disk or node failure, maintenance agents on each storage node will generate additional redundant data blocks on available storage nodes to substitute the missing data from the failed node. As the system is configured to survive multiple simultaneous node failures, this can happen in the background while data availability is never jeopardized.
If the B-Virtual DSS storage nodes are installed in multiple racks across multiple locations or data centers, this location information is used to generate data spreads as wide as possible. As such, major failures or disasters that impact full racks or even render a total data center unavailable will have no impact on data availability. In such event, the amount of lost data within a storage spread is always minimal because of this node selection policy. This minimizes the volume of data to be regenerated and the time required to regain full policy level data availability.
As data from failed storage nodes in the DSS is automatically regenerated on available nodes, there is no need to replace the failed nodes. As long as storage capacity is available, the DSS storage system will keep operating and maintaining itself without any intervention. This eliminates the need for on-site interventions and dramatically reduces the operational cost of the system.
Summarizing, a disk or full node failure in the B-Virtual Distributed Storage System does not lead to a critical situation, as the data traffic is immediately reconfigured to use available nodes. Data that was lost on the failed node is automatically regenerated on available disks in other storage nodes. This process happens in the background and has very little impact on the storage traffic. As the DSS is configured to survive multiple simultaneous node failures within the spread, a single disk or node failure does not lead to a critical situation.
The data spread strategy of the B-Virtual DSS minimizes the impact of larger rack or full data center disasters. Data is automatically regenerated on available nodes to quickly regain the data availability service level.
|