Browse Prior Art Database

Accurate resource gauging for dynamic resource management in managed systems. Disclosure Number: IPCOM000195838D
Publication Date: 2010-May-19
Document File: 1 page(s) / 53K

Publishing Venue

The Prior Art Database


Disclosed is a method for accurately gauging resources for dynamic resource management in managed systems. This helps the users in identifying the accurate resource required by each logical partition on a virtualized server over a period of time and subsequently take action to free the unused resources thus optimizing the resource usage on the server system. Virtualization can be found in a number of server management or platform management systems. These servers can utilize standard virtualization functions or logical partitioning (LPAR) technology by using system management or platform management tools. Logical partitions allow clients to run separate workloads in different partitions on the same server, potentially lowering costs and improving energy efficiency. Partitions on servers are designed to be shielded from each other to provide a high level of data security and increased application availability. Dynamic Logical Partitioning (DLPAR) allows clients to dynamically allocate numerous system resources to application partitions without rebooting, simplifying overall systems administration and workload balancing and enhancing availability. In a server management or platform management tool, the user, while assigning resources to a LPAR makes assumptions of the approximate resources required. The User tends to over-allocate the resources in order to be on the safer side so that the application does not face lack of resources during critical period. Due to over-allocation, much of the resources remain under-utilized and resource-allocation becomes in-efficient. As per the above mentioned scenario, we deduce that a solution is required in place to keep track of the unused resources and automatically notify it to the user. The following list answers to the query as to why additional solution is required for this scenario: 1. Additional solution is required to improve the efficiency of the resource usage by the micro-partitions over the time. 2. This method allows the User to know the actual resource usage by the micro-partition, which is different from the earlier assumed values. 3. During critical period, when there is scarcity of resources for some partitions, allocations can be dynamically changed based on this data to accommodate the situation. 4. This can be done without the User even knowing about it. 5. The User can also be notified of the actual resource usage over the time, so that he can dynamically de-allocate the resources or can even choose to keep it as before. Core Idea: The Core Idea of the invention is to estimate the exact resource usage of a partition on the Server over a period of time and calculating the over-allocated resources. This over- allocated resource in itself forms a pool of unused resources and can be used as the User wishes to. The solution comes up with a host of advantages for the users like: 1. It gives exact estimation of resource usage by a LPAR, compared to the values assumed by the User. 2. The values could be communicated to the User, in order to accept the statistical value, thus freeing the over-allocated values. 3. The data could be kept as a critical time back-up value, during which, based on the data acquired and the extra resources required, the System automatically dynamically de-allocates and re-allocates the unused data and accommodates the critical situation. 4. The data could be kept as a virtual unused pool, which is used during the time of crisis.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 70% of the total text.

Page 1 of 1

Accurate resource gauging for dynamic resource management in managed systems

Accurate resource gauging for dynamic resource management in managed systemsAccurate resource gauging for dynamic resource management in managed systems .



The procedure followed to implement the solution is given below:

Procedure to collect accurate resource requirement:
1. The Hypervisor monitors the resource requirement of each individual partition and estimates the accurate resource required over a period of time. The data is always kept current and updated.

2. The above data collected is compared with the initial resource requirement data assumed by the User during partition creation.

3. The difference between the user assumed data and the actual usage data is kept in a different profile which is always kept current and updated.

Procedure followed to make use of the accurate requirement data:

The data collected from above analysis can be used in a number of ways:
1. The User can be notified about the actual data requirement after a period of time and the changes which can be done to the initially assumed data. The User can either agree with the analysis data by accepting the new profile generated or reject it, in which case the earlier assumed values will still be valid.

2. During Critical times when there is resource crunch due to either damage to existing resources or due to sudden increased workload in partitions, the situation can be controlled by selectively and dynami...