Cloud Computing Models: IaaS, SaaS, PaaS

Question:

Discuss the requirement for remote administration, resource management and SLA management.
 

The IT resources in the cloud, that are shared between the end users are needed to be managed and  allocated in  a manner  such that every user can get their required resources  whenever they need and in a cost efficient manner. Now to configure, maintain and monitor the resources. All this management issues in cloud systems are maintained by using different mechanisms, such as resource management system, remote administration system, management system of SLA and payment management system.

2.1 SaaS model

In the cloud computing there are three models IaaS, SaaS, PaaS. The SaaS is the abbreviation of the software as a service. These solutions include the   email management solutions, ERP solutions, or the customer relationship management solutions.  As the services provided to the client organization through the internet, user or the client does not have to install any package in their local system. In this model the as the services from the  provider  is accessed remotely  from  the organization in the similar way the remote access also facilitates the administrator to control the access ,usage and other administrating monitoring tasks.

As an example we can take the case of ACL realty’s case study of implementing the SaaS models using the Google cloud services for managing its CRM and the email services. Google provides its own encryption and decryption methodologies to provide a extra layer of security to its cloud services .The Google provides the data store dashboard as a remote administration tool to its clients. Where the console need to have a valid user name and password to login and use the provided services. The administrator can view the data related to the applications provided by the Google cloud services. Also the statistics of composite and built in indexes can be checked from the provided data store dashboard.    

2.2 Remote administration

In the remote administration of the clouds, this mechanism provides tools like interfaces that are used by both the internal and external administrator to configure and administrate the services of the cloud. The administrative tasks are implemented on the leased resources. There are some typical tasks are to be realized or completed by using the remote administrator   interface or the console (Beloglazov et al., 2012).

1. Managing the cloud usage and the costs related to that.
2. Configuring and setting up the services of the cloud according to the business needs o the organization.

• Managing the corporate users, credentials related to security, authentication.

1. Monitoring the service level agreements and the quality of the services.
2. Tacking the services or the usage by the external or internal clients.
3. Monitoring the capacity of the cloud.

2.3 Remote administration using VPN

In case of remote administration of the cloud services using a VPN, it requires networks that can handle dual channels. Because using a VPN channel the normal user data can be  sent or received and using the another  channel  the administrative

Packets can be transmitted. The VPN helps the administrators of the corporate organizations to connect the remote branches to the   headquarters or the main office in secure, economic way. The main advantage in using the VPN in the remote administration is that, although the data channel can have a problem but the  administrator can  use the  administrative channel can reach the desired machine and work on solving the  problem. The administrations through the VPN have some problems and difficulties too. Like if some the machines are local and some of them are remote then there would be no administrator to solve the problems (Alfano et al., 2014).

Resource Management Strategies

2.4 Resource management in cloud services

In the cloud services, the resource allocation process has a very vital role that ensures the success of the total cloud services. There are different strategies adopted by the provider to manage the resources in the most efficient way. Some of those methods are as follows:

Linear scheduling strategy: In this strategy, the algorithm mainly targets to maximize the selected quality of service parameters.  The resources are allocated among the users to maximize the QoS parameters. In this strategy, the parameter is cost function (Wu et al., 2013).

Pre copy approach: In this approach, the copies of the memory are repeatedly copied from the source machine to the destination host machine. The main drawback of this approach is, it causes delay to the experience of the user (Nguyen et al.,2013).

Matchmaking and scheduling: In this approach the method of allocating the job are obtained from the resource pool and in the next step it is determined that how the requested resources will be allocated to the different processes or jobs (Erdil, 2013).

Just in time resource allocation: In this resource management technique, it is tried to minimize the idle time for any resources. To do this the time intervals are set to the resources and altering the resources with the change of workload on the clouds (Rasmi & Vivek , 2013). 

2.5 SLA

The SLA is the abbreviation is the service level agreement. It is a contract between the cloud service provider (both internal and external) and the client. It defines the expected services that a user will get from its cloud provider.

There are some guidelines that each SLA should maintain to provide the quality services to its clients.

2.6 The business cases that should map to SLA

In the organization the IT resources helps in the automation. Therefore, a misaligned SLA can cause problem in the automation process of the client organization. This can also lead to the unreasonable deviation in the  provider guaranties.

Scope of SLA: The scope of the SLA should be understood both by the provider and by the client organization. Since the cloud services are provided by integrating many  layers so it is important to know that to which extent the  guarantee applies.

Scope of the SLA monitoring: Also in case of monitoring, the agreement should ensure that where the monitoring should be done and where the performance will be measured.

Defining penalties or compensation: In case of the provider, fails to maintain the quality of service of the cloud then the provider should compensate the client. This issue also needs to be documented on the SLA.

Securing the non measurable requirements: Some client requirements cannot be measured in the usual performance metrics. Like the privacy and the data security issues. When a client is adopting a provider then the provider should maintain a level of    privacy security of the data.

2.7 Archiving the SLA data

The data acquired by monitoring the SLA should be archived. So that this data can be   presented in future for reporting purpose.

Role of SLA in Cloud Computing

For the Amazon web services the service level agreement is a policy that governs the use of its cloud services by the client under the terms of its customer agreements. As stated in its site the company tries to make the services available to the users with the uptime of 99.95% monthly. If the   provider fails to meet the commitment then the client is eligible for the  service credit depending upon some conditions that are stated by the AWS. The uptime in the service is defined by the time amount in which the cloud services provided by the provider is accessible the client or the end user.

2.8 Data backup in the cloud services

As the cloud, service providers provide a huge storage space to its clients with the higher computation power. This leads the users to store the data to the clouds, to have the access    remote locations.  This storage has the problems too.  Like if the user data gets destroyed or disappears from cloud by some administrative mistake. To avoid the loss of the data there are different    backup and disaster recovery techniques are adopted by the providers.

There are several techniques are available for the backup and disaster recovery. Some of them are as follows:

2.9 High security distribution and rake technology

This technique utilizes the mechanism process of distributed data transferring under the encryption technology. Three components are mainly used in this method, data centers, supervisory servers and client nodes. The backup sequence and the recovery sequence. When there is a need of back up then the data centers encrypts, scrambles, duplicates the data, and sends Meta data to the supervisory servers (Thomas et al., 2013). In the recovery, sequence the supervisory servers’ collects the encrypted, fragmented data using the “rake reception” procedure and then they are merged, decrypted, and as well as descrambled.

2.10 Efficient routing grounded on taxonomy

In our research, this technique is not a data backup technique but a data retrieval technique. This technique is completely dependent on the semantic comparison in service requests and descriptions. The ERGOT (AH et al., 2014) use three components i) SON (semantic overlay network) ,ii) DHT(data hash table) ,iii) measure of semantic similarity of the service request and description.

2.11 Cold and hot backup service replacement strategy      

This recovery process is activated whenever it detects the failure of the services. This technique returns the service compositions dynamically. At the time of implementation of the services, the backup services remains in the active state.

2.12 Shared backup router resources

This technique focuses on the cost reduction in the back up of the data. This technique uses the IP address that is unchanged even after the failure of the router. This technique uses the network management system to backup the data (Weiyan et al., 2014). 

3.0 Conclusion

Due to different security and the implementation flaw the cloud services provided by different providers is exposed to different security as well as loss of data threats. To avoid the loss of data due to this threats and flaws the provider as well as end user or client should be aware. By implementing the proper SLA, use of different backup and recovery techniques   can this shortcomings can be avoided, and help the organization grow by properly using the advantages of the  IT enabled  business services or cloud services.  

References

AH, N. H., Shu, G., Malek, A. G., & Zi-Long, J. (2013). An optimal semantic network-based approach for web service composition with qos.TELKOMNIKA Indonesian Journal of Electrical Engineering, 11(8), 4505-4511.

Alfano, M., Prete, D. D., Michelino, D., Pardi, S., Vela, R., & Russo, G. (2014, December). Network Management in Cloud Computing for Public Administration: A Practical Use Case. In Proceedings of the 2014 IEEE/ACM 7th International Conference on Utility and Cloud Computing (pp. 768-773). IEEE Computer Society.

Beloglazov, A., Abawajy, J., & Buyya, R. (2012). Energy-aware resource allocation heuristics for efficient management of data centers for cloud computing. Future generation computer systems, 28(5), 755-768.

da Cunha Rodrigues, G., Calheiros, R. N., de Carvalho, M. B., dos Santos, C. R. P., Granville, L. Z., Tarouco, L., & Buyya, R. (2015, July). The interplay between timeliness and scalability in cloud monitoring systems. In 2015 IEEE Symposium on Computers and Communication (ISCC) (pp. 776-781). IEEE.

Erdil, D. C. (2013). Autonomic cloud resource sharing for intercloud federations. Future Generation Computer Systems, 29(7), 1700-1708.

Nguyen, H., Shen, Z., Gu, X., Subbiah, S., & Wilkes, J. (2013, June). Agile: Elastic distributed resource scaling for infrastructure-as-a-service. In Proc. of the USENIX International Conference on Automated Computing (ICAC’13). San Jose, CA.

Rasmi, K., & Vivek, V. (2013). Resource management techniques in cloud Environment-A Brief Survey.

Thomas, G., Jose, V., & Afsar, P. (2013). Cloud computing security using encryption technique. arXiv preprint arXiv:1310.8392.

Weiyan, X., Dong, L., Ming, L., & Pengfei, J. (2014). Autonomous Recovery Technique of Software Bus Based on VxWorks Operating System. InProceedings of the World Congress on Engineering (Vol. 1).

Wu, Z., Liu, X., Ni, Z., Yuan, D., & Yang, Y. (2013). A market-oriented hierarchical scheduling strategy in cloud workflow systems. The Journal of Supercomputing, 63(1), 256-29.

Younis, M. Y. A., & Kifayat, K. (2013). Secure cloud computing for critical infrastructure: A survey. Liverpool John Moores University, United Kingdom, Tech. Rep.

Zissis, D., & Lekkas, D. (2012). Addressing cloud computing security issues. Future Generation computer systems, 28(3), 583-592.