During the active period of the client, the service guarantee for the request will be in the range between the minimum and maximum requests. Due to the presentation rate changes or buffer overflow/underflow, the client may ask the server to speed up or slow down dynamically. Thus, resources must be flexibly allocated to accommodate such changes.

Following the resource allocation approach, dynamic resource adjustment is required to handle the bandwidth changing needs of multimedia data retrieval. For instance, when a client first comes into existence, it might estimate its bandwidth requirement to be a certain value, but during the course of data transfer from the server, it might realize that this rate is in fact too low or too high.

In a scenario where the client needs to speed up or slow down its data acquisition from the server, it communicates this to the server. The server, makes certain calculations to determine the value by which the rate has to be increased or decreased. Due to differences in media data compression techniques and in the relative placement of data on the disk, the actual time h ^act_kij taken to retrieve the requested segments in round r_k could be different from h _kij. The remaining disk resources can be calculated by:
 

When a currently admitted client requests a rate increase to server s_i, if D Bⁱ_d has enough resource to allocate for the increase request, the request is then granted. Otherwise, if D Bⁱ_d does not have enough resource to allocate for the increase request, additional disk resource is unavailable and the increase request cannot be granted. In this case, advanced disk scheduling may be applied to retrieve partial data, while the presentation of the data still appears faster.