During maintaining current business partnerships or gaining a few beneficial ones, a Service-Level Agreement (SLA) plays an important role in outlining expectations and business objectives from both parties.

Rana, Warnier, Quillinan, Brazier & Cojocarasu (2018) generally specifies SLAs in his study as performance-related properties referred to as Quality of Service (QoS) terms that must be maintained by a provider during service provision.

SLA represents an agreement between IT Service Providers and their customers that contains a set of responsibilities and obligations called as Service-Level Objectives (SLO). In addition to that, SLAs also detail sets of performance that should be achieved during the service agreement.

SLA Violation

In recent years, the art and science of the Service Level Agreement (SLA) has evolved from monitoring metrics around systems and applications to a focus on business outcomes and business performance. However, when SLAs are not monitored regularly and do not align with the current business environment due to a lack of updates and revisions, they become ineffective. Additionally, SLAs become incoherent and misinterpreted, which leads to unclear messages and objectives for both IT Service Providers and their customers and causes violation. Trienekens, Bouman, and van der Zwan (1999) elaborate 3 circumstances which can trigger SLA violation as follows.

1. Unclear and Incomplete Service Specification

In certain cases, some service specifications needed more detailed information and particular facts that differentiate them from other services. According to Trienekens, et. al., a key problem is that there is not yet a good classification and there are not yet good definitions of disaster. Therefore, SLA becomes violated when the customer didn't grasp the information of service specification or when the service providers didn't specify the services they offer.

2. Insufficient Cost Management

Costs and budget of IT Services are usually fixed at the beginning of an agreement, however, there are some conditions that require stretching and that means additional costs. These conditions should be stated at the beginning of the agreement so that both parties are aware of their cost management and avoid unclear interpretation. Unclear is in that way these costs can be differentiated and can be related to specific IT Services, and in particular to the needs and the wishes of a customer (Trienekens, et. al.).

3. Unfulfilled Requirements and Obligations

SLA usually marks out details of roles and responsibilities for both parties that need to be complied with in order for the business process to run smoothly. However, without regular monitoring and evaluation, the requirements are not fulfilled and disrupt the performance results. This caused unwanted accidents or conflicts during the process.

SLA is a contract that specifies service delivery; however, SLA is sometimes overlooked and not updated. SLA, in fact, is not merely a contract, but more about expectations and standardization.

How to Resolve SLA Violation

In response to this, it is desirable for customers to detect violation to verify the quality of service they receive and potentially get their money back if there is a breach of contract (Lee, 2018). There are several research about methods that could be used to minimize SLA Violation such as low-level metrics, QoS parameters, and performance metrics using Markov Chain, Recursive Least Square, and Autoregressive Process.

The abovementioned methods can help to predict likely service violation (Hussain, Hussain, Hussain, Chang, 2016). In addition, they allow prior detection of violation and compile possible solutions to overcome them. In this article, we will discuss 3 of the methods used to minimize and resolve SLA Violation, namely Penalties, Provider-Based Optimized Personalized Viable SLA (OPV-SLA), and Cross-Layer Adaptation. Further elaboration of each method is presented in the following subsections.

1. Penalties Method

When designing SLAs, a set of penalty clauses should be included in order to determine the quality level of deliverance and ensure the commitment of both parties. It is important to lay out what type of penalty should be applied, levels of violation, from tolerable to non-tolerable. and if enforcement of the law is needed. An important issue that should be considered when designing 'penalty schemes' is that behind the imposition of any contractual sanction lies the idea that the faulty behavior of a provider should be deterred (Rana, et. al., 2018).

2. Provider-Based Optimized Personalized Viable SLA (OPV-SLA)

OPV-SLA is a methodology used to conduct a prior analysis of the customers before designing SLAs. According to Hussain, et. al., OPV-SLA presents an approach that assists the cloud provider to decide whether or not to accept a consumer's request and helps with the formation of an optimal viable SLA by offering a maximum limit of marginal resources based on the available resources. The service provider analyzes collected data with determined variables and available resources to ensure the customers' profiles which later became the core elements in designing SLAs. In this way, the SLAs will help service providers to monitor performance progress and their customers' behavior.

3. SLA Prediction and Cross-Layer Adaptation Combination

This methodology mitigates arising violation possibilities with multiple adaptation mechanisms and SLA Prediction. Vidačković, Weiner, Kett & Renner (2010) in their study presents a cross-layer monitoring and adaptation framework, where they aim at the Business Model with respect to the lower layers and propose a cross-layer adaptation strategy based on a top-down approach. The combination of multiple adaptations and SLA Prediction allows the service provider to monitor services, notify violations when occurred, and deliver negotiation strategies using available information. In that case, violations can be avoided earlier and prevented unwanted conflicts from happening.

Conslusion

SLA needs to be both prepared carefully and monitor regularly to minimize risks in the business process and gain beneficial partnerships. With suitable frameworks and approaches, violations in SLA can be avoided and mitigated effectively.

Reference:
Bouman, J., Trienekens, J., & van der Zan, M. (1999) Specification of Service Level Agrements, Clarifying Concepts on The Basis of Practical Research. DOI:10.1109/STEP.1999.798790
Hussain, W., Hussain, F. K., Hussain, O. K., & Chang, E. (2016). Provider-Based Optimized Personalized Viable SLA (OPV-SLA) Framework to Prevent SLA Violation. The Computer Journal, Volume 59, Issue 12, 1 December 2016, Pages 1760–1783, https://doi.org/10.1093/comjnl/bxw026