Availability Management in Cloud Computing

Pre-requisite: Cloud Computing and Cloud Based Services

Cloud Services are not immune to outages (failure/interruption) and the severity and the scope of impact on the customer can vary based on the situation. As it will depend on the criticality of the cloud application and its relationship to internal business processes.

1. Impact on business: In the case of business-critical applications where businesses rely on the continuous availability of service, even a few minutes of service failure can have a serious impact on the organization’s productivity, revenue, customer satisfaction, and service-level compliance.
2. Impact on customers: During a cloud service disruption, affected customers will not be able to access the cloud service and in some cases may suffer degraded performance or user experience. For Example:- when a storage service is disrupted, it will affect the availability and performance of a computing service that depends on the storage service.

For example, on December 20, 2005, Salesforce.com (the on-demand customer relationship management service) said it suffered from a system outage that prevented users from accessing the system during business hours. Users “experienced intermittent access” because of a database cluster error in one of the company’s four global network nodes, company officials said in a statement the day following the outage.

Factors Affecting Availability:

The cloud service’s ability to recover from an outage situation and availability depends on a few factors, including the cloud service provider’s data center architecture, application architecture, hosting location redundancy, diversity of Internet service providers (ISPs), and data storage architecture. 

Following is a list of the major factors:

• The redundant design of System as a Service and Platform as a Service application.
• The architecture of the Cloud service data center should be fault-tolerant.
• Having better Network connectivity and geography can resist disaster in most cases.
• Customers of the cloud service should quickly respond to outages with the support team of the Cloud Service Provider.
• Sometimes the outage affects only a specific region or area of cloud services, so it is difficult in those cases to troubleshoot the situation.
• There should be reliability in the software and hardware used in delivering cloud services.
• The infrastructure of the network should be efficient and should be able to cope-up with DDoS(distributed denial of service ) attacks on the cloud service.
• Not having proper security against internal and external threats, e.g., privileged users abusing privileges.
• Regular testing and maintenance of the cloud infrastructure and applications can help identify and fix issues before they cause downtime.
• Proper capacity planning is essential to ensure that the cloud service can handle peak traffic and usage without becoming overloaded.
• Adequate backups and disaster recovery plans can help minimize the impact of outages or data loss incidents.
• Monitoring tools and alerts can help detect and respond to issues quickly, reducing downtime and improving overall availability.
• Ensuring compliance with industry standards and regulations can help minimize the risk of security breaches and downtime due to compliance issues.
• Continuous updates and patches to the cloud infrastructure and applications can help address vulnerabilities and improve overall security and availability.
• Transparency and communication with customers during outages can help manage expectations and maintain trust in the cloud service provider.

System as a Service Customer’s Responsibility:

• Customers should understand the Service Level Agreement(SLA) and communication methods so that they will be informed on service outages or maintenance.
• Customers should be aware of options to support availability management that is they should understand the factors affecting availability management.
• The customer of System as a service should be aware that the cloud service is multitenant which means Cloud Service Providers typically offer a Standard Service Level Agreement(SLA) for all customers. Thus, Cloud Service Providers may not be able to provide their services to the customers if the standard Service level-Agreement(SLA) does not meet the service requirements. However, if you are a medium or large enterprise with a big budget, a custom SLA can be made available.
• The customers should be aware of how resource democratization occurs within the Cloud Service Providers to best predict the likelihood of system availability and performance during business fluctuations.
• Customers should ensure that their applications are designed and deployed in a way that maximizes availability and resilience. This may include using load balancing, redundancy, and failover mechanisms.
• It’s important for customers to monitor their own applications and infrastructure to detect and respond to issues quickly, rather than relying solely on the cloud service provider to do so.
• Customers should understand the security and compliance implications of using a cloud service and take appropriate measures to protect their data and systems.
• It’s important for customers to have a disaster recovery plan in place, including backups and a procedure for restoring service in the event of an outage.
• Customers should understand the cost implications of using a cloud service, including any charges for exceeding usage limits or for premium support options.
• It’s important for customers to provide feedback to the cloud service provider on their experience using the service, including any issues or suggestions for improvement.
• Customers should understand the limitations and restrictions of their cloud service subscription, such as the maximum number of users or the amount of data that can be stored, and plan accordingly.

System as a Service Health Monitoring:

The following options are available to customers to stay informed on the health of their service:

• Service dashboards should be published by the Cloud Service Providers So that they can publish the current state of services and can also inform the outage or any kind of maintenance of the cloud.
• Customer should check their mailing list as the service provider might have notified them about recently occurring outrages.
• Use third-party tools to check the health of the application.

Platform as a Services Customer’s Responsibilities:

The following considerations are for Platform as a Services Customers:

• PaaS platform service levels: Customers should read and understand the terms and conditions of the Cloud Service Provider’s Service Level Agreements.
• Third-party web services provider service levels: When your Platform as a Services application depends on a third-party service it is critical to understand the Service Level Agreements of that service.  Network connectivity parameters with third-party service providers. Example: Bandwidth and latency factors.
• Platform as a Service Health Monitoring: The following options are available to customers to monitor the health of their service:
  • Service health dashboard published by the Cloud Service Provider.
  • Cloud Service Providers customer mailing list that notifies customers of occurring and recently occurred outages
  • Use third-party tools to check the health of the application
• Infrastructure as a Service Health Monitoring: The following options are available to Infrastructure as a Service customer for managing the health of their service:
  • Service health dashboard published by the Cloud Service Providers.
  • Cloud Service Providers customer mailing list that notifies customers of occurring and recently occurred outages.
  • Third-party-based service monitoring tools that periodically check the health of your Infrastructure as a Service virtual server.