Tag Archives: cloud services

Cloud and the role of SLA

As cloud computing is taking center stage for different IT enabled business enterprises it is highly essential to define policies, procedures and service level agreements (SLA) in order to maximize the value of cloud for both the consumer and the service provider. SLA statements written must be measurable, achievable, relevant and timely and should remain specific for cloud services aimed at minimizing ambiguities for both the cloud consumer and the cloud service providers.

The cloud service models (IaaS, PaaS, SaaS, etc.) offer new paradigms of computing resources and IT enabled capabilities for all types of organizations. IT industry experts claim that over 80% of enterprises have adopted some cloud service in their organization. The key term ‘service’ in cloud computing creates the need to develop contracts named service level agreements (SLA) between the client organization and the cloud service provider (CSP). SLAs are used by companies for a long time, especially when the company hires third party service provider to manage some of their business operations. SLAs will ensure the consumer receives all the services availed as agreed by the provider and of course ensure money’s worth for the client.

Likewise, an organization deciding to hire cloud services for their IT needs, SLAs come into play to make sure the services offered by the CSP are delivered as promised. SLA has become a pre-requisite due to cloud business strategy and provides series of rules and directives that must be taken by cloud consumers to evaluate and negotiate terms with CSP. It describes a set of non-functional requirements of cloud services. An example of SLA can be the return of operations (RTO) in case of any service failure in the cloud.

Cloud SLA is imperative for compelling reasons,

  • Ensure availability and uptime
  • Specific performance benchmarks to compare actual cloud performance
  • Availability of usage statistics for the consumer
  • Informing scheduled changes to consumers in advance (eg., maintenance downtimes)
  • Help desk and support to resolve specific issues
  • To clarify the scope of resources used in cloud service of interest

SLAs are the means of documenting cloud services between the CSP and consumer and play a major role for the following reasons:

    • Roles and Responsibilities: Consumers must understand the roles and responsibilities and business relationships between them and the CSP. For example, an indirect actor namely cloud carrier is an entity providing the carrier or transport for cloud services between CSP the consumer. In this scenario, the SLA must cover provisioning of alternative carrier in case of non-availability our outage with one carrier. According to NIST (National Institute of Standards and Technology) reference architecture, the actors involved in cloud are: Consumer, CSP, Auditor, Broker and Carrier, with unique roles. Cloud consumers must recognize and understand the activities and roles of each entity or service in the cloud as explained by CSP including their own set of responsibilities.
    • Examine Business Level Policies: Business level SLAs would define Guarantees provided by the CSP (for example, guarantees will include 99.99% uptime, measurable performance and usage, etc.). Acceptable use policy is a business level SLA statement where the CSP describes how the service should be used, List of services not covered and Excess usage. Normally, the CSP will encourage the consumer to buy resources that is only required for their business. Other policies will include Payment and penalty models, Activation, Renewals, Transferability, Sub-contracted services, Licensed Software, Industry specific standards and Support.
    • Data Level Policies: Data level policies are critical in SLA. Here CSP will explain on how the consumer’s data is governed and protected in local jurisdiction or other locations where the data will reside or made available. Consumers must carefully evaluate legal requirements on how SLA will handle issues related to movement of data to offer multi-site storage in different jurisdictions for redundancy. The other critical SLAs in data level policies include, Data Preservation – backup, restore, redundancy, etc.; Data Locations – will verify data locations for consumers; Data Privacy – defines how consumer data is secured and used; Data Seizure – in some circumstances the data can be seized by government agencies, etc. Therefore, data level policies in SLA are the most critical policies which must be evaluated thoroughly by consumers.
    • Service and Deployment Model Differences: Service models are categorized as IaaS, PaaS and SaaS. The service models in cloud are unique in terms of service delivery and are defined with unique SLAs. Likewise cloud deployment models are private, public and hybrid clouds which have a unique set of SLAs. According to Cloud Standards Customer Council (CSCC), consumers should understand the nuances of service and deployment models and their corresponding SLAs because their value and risk varies significantly.
    • Describe Objectives for Critical Performance: SLA in performance objective relates to efficiency, accuracy and service delivery. Performance statements in the SLA will help consumers to measure and audit different aspects on cloud performance. Performance metrics are dependent for each service IaaS, PaaS and SaaS. For example, performance considerations for IaaS will include network and compute and so on.
    • Security and Privacy Considerations: SLAs related to security and privacy considerations deals with information assets – data, applications, functions and processes and can be defined based on criticality and sensitivity of consumer data. Normally CSPs offer global security standards defined in standards such as ISO, COBIT, ITIL, etc. The SLA will also cover alternative actions in case of security breaches or data loss for the consumer.

In addition to the above roles, SLAs will also define areas such as disaster recovery, service management, auditing, self-service metering and provisioning, solutions for service failure, remedies and limitations in cloud services. SLAs will also state exit processes followed in case a consumer wishes to discontinue from a service provider.

Data Center Infrastructure is the key to Cloud Service Availability

Cloud computing services are basically provisioned in data centers. Organisations while planning to adopt cloud computing systems must ensure to evaluate the data center which provides the underling foundation and infrastructures for a cloud service model. Tier 4 standard data centers are ideal for clouds because they provide adequate redundancy and fault tolerance to ensure service uptime and zero data loss which is important for cloud service availability.

Widespread interest in cloud computing systems among IT circles is fostering the evolution of different infrastructure models, service levels, and application platforms. Basically cloud computing is a model or infrastructure that provisions resources dynamically and makes them available as services over the internet. Cloud models and services can be accessed by organisations by availing resources as appropriate to satisfy their IT needs. Cloud service levels include Infrastructure as a service (IaaS), Platform as a service (PaaS) and Software as a service (SaaS). Cloud computing models are classified as three major types’ namely private clouds, public clouds and hybrid clouds.

Cloud service infrastructures have the ability to scale up or down in terms of storage, compute, and network depending on usage. Organisations view the cloud as a viable alternative to outsource their entire IT infrastructure primarily due to cost savings in terms of capital and operational expenditure. Many organisations have adopted CtrlS cloud model(s) in their business operations and are able to realize business advantage. Cloud computing services are available with Data Centers (DC) in different variants to fulfill organizational business needs. An organisation planning for adopting cloud computing model for their IT operations must ensure to critically understand the architecture of DC from where cloud model is availed.

The key criteria for evaluation include:

  • The underlying infrastructure
  • Redundancies
  • Connectivity
  • Uptime
  • SLAs
  • Services
  • These six areas must be considered by organisations from the perspective of cloud capabilities for continuous service availability and reliability. The above criteria are explained in a bit more detail to provide further understanding on existing cloud capabilities available with CtrlS Data Centers.The underlying infrastructure of DC plays an important role in cloud computing because when business grows the infrastructure has to accommodate increasing volumes of data and network traffic. The increasing surge in online transactions both by users and businesses require a cloud infrastructure to handle high volume data growth in terms of storage, network and processing. Most organisations have stringent policies on data privacy. CtrlS Data Centers offers private cloud models for businesses and organisations. Private cloud models offer fully secured data network with robust security features to protect information from falling into wrong hands. The cloud infrastructure is fully flexible, it allows users (organisations) to provision and monitor their resources themselves with automated metering service. CtrlS DC infrastructure is fully equipped to handle high volumes of data and perform complex processing on them without affecting efficiency and performance for users.Redundancy in DC is a measure of fault tolerance and also classifies the data center standard. DC standards are basically classified between Tier 1 to Tier 4 standards depending on their power capacities and their ability to quickly restore operations in the event of outage or in the event of component failure. CtrlS data center operations handle lot of mission critical data used by enterprise organizations. Hence the infrastructure cannot afford to experience any type of systems failure. CtrlS DC is built to fulfill Tier 4 standard where each component in the data center is provided with adequate backup components thus ensuring zero outage. Tier 4 data centers have the ability to respond automatically to any kind of failure or hardware malfunction. Hence failure on any one component does not affect other components in the infrastructure thus ensuring zero data loss and zero downtime. Organisations desiring to adopt cloud computing must check for redundancy in DC which has an effect on cloud service availability. Availability is a critical business need.

    Connectivity refers to the network connections on how the components are connected with each other to send and receive data in the DC. DCs are driven by consolidation, virtualization and adaptive architectures that enable flexible visions such as IT as a service, cloud computing and autonomic computing. I/O is a key differentiator because connectivity encompasses many components such as switches, routers, gateways and backbone networks and is measured in terms of throughput and bandwidth. CtrlS cloud models are virtualized to support Ethernet and Fiber channels while each virtual machine (VM) is viewed independently for its configuration. VMs are stored in centralized storage which separates the VM from a single physical server. In order to support multiple VMs in the data center the network infrastructure must provide reliable, flexible and secure data transport. CtrlS data centers are equipped with switching architectures to deliver low latency, throughput and high speed inter connects to support cloud models for efficient transport between servers, storage and end users.

    CtrlS DC is built on Tier 4 standards to ensure fault tolerance and provide guaranteed 99.99% data center uptime. This is an important Service Level Agreement (SLA) metric in terms of service availability and quality of service guarantee. SLAs are defined for the services available in cloud computing models. The user (organisation) may choose the right amount of resources in the cloud and adhere to the corresponding SLAs that define with each resource. For example, SLAs are defined specifically for resources such as storage space, backup, network bandwidth, number of VMs used, etc. SLAs that define security, data privacy, firewalls, uptime, etc are common to all users. The organisation must ensure to define SLAs to be specific, measurable, achievable and timely and most importantly there must be no room for any ambiguity for both cloud service provider and users. CtrlS SLAs are defined to revolve around performance metrics that meet business needs for users availing cloud services.

    CtrlS DC provide additional services and features that include different levels of security, firewalls and intrusion detection systems to mitigate any types of attacks and intrusions. This ensures the entire cloud model fully secure. Further users availing cloud models from CtrlS benefit from a wide range of services such as automated monitoring, UI for resource provisioning, zero downtime, zero CapEx (capital expenditure), etc. Tier 4 standards provide the much needed fault tolerance to ensure high availability at all times.

    Data centers are the foundation for clouds. They provide the underlying infrastructure for clouds to run. Data centers are built with thousands of servers, storage disks, switches, routers and other components to provide security, virtualization and application resilience. In a cloud models users normally look for scalability, reliability and service availability at all times. CtrlS cloud is available for enterprises either big or small. Organisations availing the cloud can benefit immensely in terms of speed and efficiency to generate business value.

Manage your Private Clouds with Full Scale Automation

Organisations adopt private clouds to eliminate issues in security and data privacy. Private clouds offer immense benefits in terms of reliability, availability and scalability. Private cloud offerings from CtrlS Data Centers provide the best in class cloud infrastructures that can scale up or down according to varying application loads and operate your business without disruptions and latency. Full scale automation is a set of intelligent tools in our cloud infrastructures meant to provide efficiency in process cycles thus minimizing human intervention in fast changing application loads. Private cloud infrastructures in CtrlS are fully automated to simplify management tasks for administrators and users alike.

Private clouds are receiving much greater attention nowadays due to its unique benefits in narrowing the growing gap between business needs and in-house IT capabilities. Cloud computing systems offer a consolidated platform with immense benefits in terms of compute, storage, networks and servers for enterprises looking to outsource their IT infrastructure. Cloud solutions from CtrlS offer scalable private cloud platforms for enterprises to take care of their growing demands in sharing compute resources across multiple applications dynamically as compared to traditional environments which are static. Automated provisioning, orchestration and capabilities in self-service are the key characteristics of private clouds to address operational challenges and to handle automation at scale. In order to overcome virtualization challenges in complex IT environments, full scale automation from CtrlS is the most ideal solution for enterprises planning on private cloud deployments.

Given below are key private cloud strategic differentiators from CtrlS in relation to business agility:

1. Standardization : Standardization refers to the efficient use, management and maintenance of technology resources. Open cloud standards are available for enterprises by which they can move on or off easily.

2. Business agility : Agility in a private cloud accelerates the creation and provision of services required by enterprises. Business applications experience peak loads in their operations at certain times. Private clouds must allow business users to provision resources by themselves independently to handle resource requirements to use their applications while restricting access to authorized application owners.

3. Superior performance and resiliency : This feature ensures reliability and service availability in private clouds. Dynamic resource allocation strategy offers resilience and superior performance.

4. Resource Optimization : Leverages existing resources for optimal use in order to reduce costs.
Full scale automation in a private cloud transforms application resources, provisioning and de-provisioning of workloads, and also helps in migrations. In manual systems these activities would require days or weeks to accomplish, but now they can be accomplished within minutes. Service menus and standardized templates are readily available to reduce human error. Cloud automation from CtrlS also transforms SLAs and end-to-end performance by way of policy driven provisioning and built in analytics thus setting the stage for self-service capabilities. Recognizing the fact that many enterprises are adopting private clouds, CtrlS offerings on private clouds are intended to increase IT agility, maximize employee productivity, accelerate business and IT alignment, optimise resource utilization and reduce costs.

Salient Features of Full Scale Automation Differentiator

In CtrlS, multi-tiered IT services and support are delivered by NOC (Network Operations Center) help desks. Normally data centers while developing solutions for managing incidents respond to problems rather than to prevent them proactively. Some of the striking features of full sale automation services include,

• NOC Support : NOC support system automates incident ticket processing and prioritizing it for immediate response. Proactive problem prevention is one of the core functions in IT service delivery. This support is ensured in private cloud systems.

• Incident Lifecycle : When an incident occurs the processes are automated for resolution and classified for its severity with quality audit. It is highly important to note that incidents are often indicators of bigger problems or underlying errors. Hence any incident that may arise is handled with high priority to prevent further errors and problems.

• Customer Portal Data Flow : Data flow through servers and systems for voice data and network services are managed and measured for communications providers. Customer portal is automated to bring all functions under technical and non-technical areas for fully convergent billing solutions based on different types of service usage.

• Automated Provisioning for Cloud : Automated provisioning enables users to configure resources as per their application needs.

• Automated metering and Billing : Usage reports to clients and Admin is automatically sent with billing information based on resource usage. The user friendly interface allows better control on billing / metering.

• Service Automation and Alerting : Services are automated for load balancing, cluster and firewall management. Incidents or errors are sent through email or SMS to ensure timely intervention. Alerting is also automated.

• Automated Restful APIs : Restful APIs are a set of principles used to design web services that focus on usage of system’s resources such as how the state of a resource is addressed and the status is sent via HTTP to wide range of clients. This entire process is automated in the private cloud systems.

CtrlS Data Centers innovate constantly to improve the existing infrastructure with new features and services. To illustrate an innovation, in addition to the features explained above our data centers have planned to implement learning management system (LMS) solutions in the cloud and auto-correction and correlation engine to generate alerts sent by the monitoring team. The LMS shall provide training portal for real time scenarios supported by centers of excellence (CoE).

Full scale automation for private cloud systems in data centers is the most viable option for enterprises to overcome issues in reliability, protection, compliance, SLA and governance. Through full scale automation CtrlS attempts to overcome challenges in traditional data centers and offers full service capabilities by eliminating manual intervention in operations.