Collaborative Cloud computing is an emergent and encouraging computational prototype for sharing on-demand cloud resources in a multi-cloud environment. In a collaborative cloud environment, we make use of unused computational resources within an interconnected network to provide seamless service to the customers. SLA (Service Level Agreement) means a legal agreement signed among the consumer and the cloud service provider to guarantee a better quality of service. Nowadays due to more user request, suppliers of cloud service are not able to deliver guaranteed Quality of Service (QoS) to the customers leading to SLA violation. Existing SLA- based resource allocation models in today's cloud market consider an agreement for each QoS parameters independently and follow fixed pricing models, which results in lacking an end-to-end approach for the client task that requires composing multiple services and inefficient utilization of computing resources, which has direct negative effect on performance and cost. In this paper, we have designed a framework model for the effective dynamic provision of resources considering various categories of SLAs laterally with their penalty. We used Hybrid Cuckoo Search algorithm for scheduling of user requests in collaborative cloud computing environment. The proposed framework services the customer requests from its interconnected cloud servers and has an SLA Negotiation policy to minimize the penalty cost and maximize customer satisfaction by reducing SLA violations.