In a plain and understandable desire for achieving economies of scale, a cloud orchestration software system should be capable of managing a huge farm of hardware servers. However, even using most advanced software configuration and management tools, the field has trial-and-error reached a common sense that the distribution scale of a cloud orchestrator mustn’t be too large. E.g., VMware, probably among the most experienced players in the trade, stipulates a rule-of-thumb upper bound for its orchestrator vRealize: No more than 1,000 servers per vRealize even if the software is installed on top quality hardware devices. Scaling-up beyond that level, cloud operation and maintenance would become unstable plus incur sharp cost increasements in operation and maintenance. Recent achievements in hyper efficient CPU virtualization by Docker have seminally ignited additional orders-of-magnitude explosions in the number of micro-servicing CPUs, certainly to add further troubles to worsening scalability in cloud orchestration. Current poor scalability status quo in cloud orchestration means that today’s clouds are in small isolated scatters and patches, and therefore cannot efficiently tap cloud potentials from economies of scale.
The essential problem behind poor scalability in cloud orchestration is that all cloud orchestrators, from commercial offerings or from open source projects, unanimously and conventionally evolve from a horizontally tight coupled architecture. A horizontally tight coupled orchestrator is a bunch of software components which are host knowledge interwoven. By speaking of "host knowledge interwoven", we mean that the software components in a cloud orchestrator know the existence, roles and duties of one another right at their birthday of being installed on a farm of server hosts, and throughout their remainder entire lifecycles afterwards. When a farm gets large, some queues of events and messages will inevitably become long; writelock mechanisms for consistency protection and CoW DB accesses will also aggregate momentum to slow down responsiveness; and occasional unfortunate popup of failures, even merely in a benign timeout sense, occurring at one point in the farm would highly likely pull down other knowledge interwoven parts. As a matter of fact, all cloud servicing or hosting providers, as long as having a size, all have to rely on human based operation/maintenance teams 7x24 on-guard the farm, playing similar roles of firefighters!
DaoliCloud presents Network Virtualization Infrastructure (NVI) technology to horizontally decouple cloud orchestration. The NVI technology minimizes the size of a cloud orchestration region down to over one single hardware server, e.g., in the formulation of OpenStack all-in-one installation. An orchestrator managing only one server host of course has absolutely no knowledge whatsoever about any other orchestrator managing another server. Thus, any server host in an NVI farm has no software knowledge about any other server host in the farm. While having obviously maximized stability for cloud operation and maintenance, the overlay cloud resources which are pooled by NVI remain to have unbound scalability. This is because NVI can trans-orchestrator connect overlay nodes in user mode only upon one node initiating communication to another (think of http connection!). NVI can connect various virtual CPUs over independent and heterogeneous cloud orchestrators, e.g., connect lightweight micro-servicing Docker containers and heavy-duty hypervisor VMs, which are independently orchestrated by, e.g., Kubernetes and OpenStack. Moreover, NVI can transparently link different cloud service providers, also in user mode.
The key enabler for any two not-knowing-one-another orchestrators to serve user-mode connection for their respectively orchestrated overlay nodes is a novel OpenFlow formulation for forwarding trans-orchestrator underlay packets. This new SDN formulation succeeds constructing any OSI layer, any form of overlay network without any need of packet encapsulation, i.e., without using any of the trans-host-network protocols such as VLAN, VXLAN, VPN, MPLS, GRE, NVGRE, LISP, STT, Geneve, or any such we have missed from the enumeration! Having avoided trans-host packet encapsulation, there is of course no need for the involving orchestrators to know one another in host mode, neither in the system installation time nor in their remainder entire lifecycles afterwards. It is in such a simple principle that the SDN innovation of NVI achieves complete horizontal decoupling of cloud orchestration. With connection taking place only in user mode, cloud deployment, operation, maintenance, and system upgrading, etc., can become 100% automated. It is now also plainly manifested that the NVI technology supports inter-cloud patching, also in user mode.
With the problem-solving architecture of NVI for truly scalable cloud orchestration, DaoliCloud attempts to contribute to the cloud industry a new production line: “Build, ship and low-cost operate any cloud, any scale”, as a new frontier to work with, extending from the great inspiration of “Build, ship and run any app, anywhere” from Docker. Having fixed a single small size for orchestrator installation and configuration, build, ship, operate and maintain any cloud, private or public, will be low cost, and fast because of one size and automation.
URL http://www.daolicloud.com exposits, in "for dummies" simplicity, a near-product-quality prototype of our new cloud orchestration technology which horizontally decouples globally located orchestrators. These globally distributed, host-mode not knowing each other, and user-mode being well organized orchestrators are independent all-in-one OpenStack hosts located in Beijing and Shanghai, China, and Virginia, USA. We cordially invite much respected reviewers of this abstract, and hopefully many subsequent interested trial users in the audience of the forthcoming OpenStack Summit, to signup on the above URL for a trial use. We humbly wish that some trial users might come to an appreciation that the new architecture work for scalable cloud orchestration can indeed enable a number of never-known-before useful cloud properties which are probably only made possible from the new architectural innovation in cloud orchestration as a specific application, and in network virtualization for a more general pursuit in knowledge and technology advances.