Sunday, 5 November 2017

Cloud Hosting Providers

The popular platforms in use are AWS, Microsoft’s Azure and Google Cloud Platform. Besides those operated by software majors IBM, SAP, and Oracle. The companies have massive investments in data centers available globally. 

They offer storage space on the pay-as-you-use basis to companies opting for cloud computing. The public cloud gives real value proposition for enterprises at low-cost and easy-to-use infrastructure. The cloud hosting grows with the business. The cloud hosting is attracting new startups and also companies already having hosting services are moving towards cloud hosting. 

On the secured environment of Cloud Hosting, companies host websites, mobile applications, and Android apps.

Furthermore, the cloud provides security, control, and performance with:

  • Full flexibility and customization.
  • High availability and reliable delivery
  • Advanced monitoring capabilities
  • Hosting Solutions offered on Cloud include cloud and hosting solutions, managed services, enterprise application management, professional services

On the Cloud Hosting, most customers are worried about the cost incurred due to the traffic peaks. Getting the high-traffic is fine for some customers, but for few, they need Throttle policy option. It could impose a limit on the volume per period, which exceeded the request incurs a counter-based delay penalty or refused. 

First, whenever the elapsed time exceeds the period length, then the volume and elapsed time are halved. Second, if the volume is below the limit, then the delay counter is decreased by one second if it is not yet zero. Otherwise, when the limit is exceeded, the delay counter is increased by one second. The delay can be between zero and ThrottleMaxDelay seconds, after which the request will be refused to avoid occupying servers unnecessarily.

The other throttle policies you can use are Random, Request, Speed, and Volume:-

Random randomly accept a percentage (limit) of the requests. If the percentage is zero (0), then every request is refused; if the percentage is 100, then all requests are accepted. The period specifies how long data is accumulated before the counters are reset.

Request imposes a limit on the number of requests per period. When this limit is exceeded all further requests are refused until the elapsed time exceeds the period length, at which point the elapsed time and counters are reset.

Speed imposes a limit on the volume (kbytes sent) per period, which when exceeded the request incurs a calculated delay penalty or is refused. First, whenever the elapsed time exceeds the period length, then the limit (allowance) is deducted from the volume, which cannot be a negative result; also the period length is deducted from the elapsed time. Second, if the volume is below the limit, in which case the request proceeds without delay. Otherwise, the request is delayed between one and ThrottleMaxDelay seconds. If the delay would exceed ThrottleMaxDelay, you refuse the request entirely to avoid occupying servers unnecessarily. The delay is computed as one plus the integer result of the volume times 10 divided by the limit.

Volume imposes a limit on the volume (kbytes sent) per period. When this limit is exceeded all further requests are refused, until the end of the period at which point the elapsed time and counters are reset.

No comments:

Post a Comment