Decenter Storage Engine
Decenter Storage Engine
High-Performance Decentralized Virtual Machine Storage_
[ TYPE : Built-in ]
Syborg deorchestrates BTRFS [B-Tree Filesystem] to provide an All-Automatic - High-Performance Storage provisioning for Decentralized Desktop and Server Virtual Machines.
Storage is now transparent and invisible with near-zero administration and management.
External storage from multiple different types of storage back-ends can also be added on-demand by directly importing/mounting them into the Decentralized Virtual Machine.
For example, Distributed Block Storage [CEPH etc], Distributed Filesystem [GlusterFS etc], iSCSI, NFS, FCoE [Fibre Channel over Ethernet], AoE [ATA over Ethernet] and others can all be directly imported/mounted into the Decentralized Virtual Machine.
The imports/mounts can be individual or a combination of multiple different types of storage backends [eg. NFS, FCoE, iSCSI etc] depending on the cost and performance requirements, and depending on what might already be in place.
When the Decentralized Virtual Machine relocates [migrates], then it will only carry forward its own thin-provisioned virtual disk-based storage along with the application[s] if installed on it, while still maintaining seamless connectivity to its imported/mounted external storage. This vSoC flexibility allows for extreme cost and performance optimizations and gains for the type & requirements of the application[s] running in the Decentralized Virtual Machine[s].
Replica_
[ TYPE : Built-in ]
The SynchroKnot vSoC Replica is a Fully-Independent, Thin-Provisioned, Writable Virtual Machine Snapshot feature with easy and flexible Time Travel capabilities.
░█ Time Travel with Spatial Virtual Machine :
Move back and forth in time with specific granularity to the preferred state of the virtual machine to the point in time when the replica was created. Multiple simple and complex scenarios are possible.░█ Time Travel with a Replica is very flexible:
Other virtual machine snapshot solutions have varying degrees of inflexibilities and complexities, and require careful management.Here are few examples of other snapshot solutions:
- The main image from which a snapshot is created cannot be touched, if not, the snapshot images get corrupted.
- The main image cannot be deleted if a snapshot exits, if not, it corrupts the snapshot images.
- In some more advanced snapshot systems, the snapshot and main image may be altered, updated etc, but to remove the main image, the snapshots need to be removed first.
Also in the cases where there are a chain of snapshots i.e Main Image --> Snapshot 1 --> Snapshot 2 --> Snapshot 3 ... to delete Snapshot 1, Snapshot 2 and Snapshot 3 may be required to be removed first.
░█ With the vSoC Replica feature, there is zero dependence between the snapshot[s] or chain of shapshot[s] and their main image. So, you can time travel as needed *without* having to worry about any sort of snapshot dependencies and inter-dependencies!
░█ Plus, making a new Replica of a virtual machine is instantaneous, and the virtual machine can be in a running state, paused or stopped state!
░█ A Replica can be created from an existing virtual machine or a spacesuit which may be in a stopped, paused or running state *without* any interruption or disruption to the existing spacesuit/virtual machine.
FASTR - Fast Asynchronous Triggered Replication, Recovery & Disaster Recovery_
[ TYPE : Power Module ]
FASTR is a per-Decentralized Virtual Machine variable block-size [ very large-to-small ], fast, intelligent, asynchronous replication.
Only triggers asynchronous replication when a block is written to inside a virtual machine. [ no agent / software needs to be installed inside the virtual machine(s) ]
Features:
▸ Frequency of the replication can be customized by setting a time interval.
▸ Priority [ Normal, High, Very High, Low, Very Low ] of the replication can be set to a custom level.
▸ Bandwith of the replication can be shaped to a custom level.
▸ Block-size of the replication data transfer can be defined to a custom size.
▸ Recovery to the original virtual machine from its replicated site can be accomplished with a simple vSoC Trigger.
▸ Recovery to the original virtual machine is also possible at its replicated site by simply creating a new virtual machine from the replicated one.
▸ Allows for very fast transfer of changed blocks of the virtual machine by building a substantially small list of checksums [ due to large blocks being checksummed ] but transferring more data inturn [ which is not a problem within and across most data centers ].
▸ Allows for slower transfer to destination with links with less bandwidth and high link latency by lowering the block-size.
Synchrosync_
[ TYPE : Power Module ]
Synchrosync is a variable block-size, fast, intelligent, asynchronous decentralized replication of virtual machine disks, spacesuits and other metadata from any specific vSoC to another specific vSoC at the same or globally different location.
Features:
▸ Replication status can be displayed with full information about the replication[s] and their progress along with the ability to stop any running replication with a click of a button.
▸ Priority [ Normal, High, Very High, Low, Very Low ] of the replication can be set to a custom level.
▸ Bandwith of the replication can be shaped to a custom level.
▸ Block-size of the replication data transfer can be defined to a custom size.
Storage Datapath_
[ TYPE : Built-in ]
The vSoC Storage Datapath allows for unique parallel push and pull of data [virtual machine disks, ISOs, spacesuits] to and from any vSoC[s] situated locally or globally.
The parallel asynchronous pushed or pulled data can be of variable block-size and its priority can be set and its bandwidth shaped to a custom level.
Storage Datapath can also be used as a ☼ unique pull-based back-up and/or archiving system ☼.
An extremely flexible and powerful capability of the Storage Datapath is the ease with which one can perform parallel secure, controlled updating and versioning of Decentralized Virtual Machine Spacesuits [the virtual machines from which other virtual machines are created].
☼ For example, the main or original copy of the virtual machine spacesuit[s] can be kept at your chosen location[s] [ eg. control center, lab, head office, etc.], then you can update those virtual machine spacesuit[s] with security, package and operating system updates/upgrades, test the operational functionality end-to-end, and then push in parallel only the changed data blocks to as many vSoCs that have those virtual machine spacesuit[s]!!! ☼
SynchroKnot Limited, Hong Kong, SAR of China [ website - content ] by SynchroKnot is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Based on a work by its creator Mehul Sharma at SynchroKnot Limited, Hong Kong, SAR of China : synchroknot.[com|hk|cloud|org].