International Business Machines Corporation

Stacy Renee Joines

Ruthie D Lyle

Kyle Gene Brown

Brian Ward Thomson

Christopher Henry Gerken

Robert M H Dunn

James M Stephens

Martin J Presler-Marshall

Ruth Edith Willenborg

Lawrence E Clark

Bruce Walter Hanover

In an illustrative embodiment of a method for managing a resource pool for deadlock avoidance, a computer receives a request from a thread for a connection from the resource pool, and determines whether the thread currently has at least one connection from the resource pool. Responsive to a determination that the thread currently has at least one connection from the resource pool, a new concurrent connection from one of a reserved partition of the resource pool is allocated and the connection is returned to the thread.

Managing resource pools for deadlock avoidance

A computer implemented method, an apparatus, and a computer usable program product for controlling startup request flooding. An application server instance is started. If the application server instance is throttled during a startup interval, a Web container thread manager is initiated. If the Web container thread manager identifies it is time to increment a Web container maximum thread pool size for the application server instance, the application server instance is notified to increment its Web container maximum thread pool size by a programmable amount. If the Web container maximum thread pool size for the application server instance has not reached a final maximum thread pool size, and if the startup interval for the application server instance has not completed, the Web container thread manager is suspended for an interval, after which the Web container thread manager identifies if it is time to increment the Web container maximum thread pool size for the application server instance.

Incremental web container growth to control startup request flooding

A servlet filter receives all incoming requests for dynamic content and can be used to ensure that the load on a server does not become too great during the time the server is engaged in start-up. The filter is configured to be the first filter encountered and will evaluate a series of conditions to determine if the request is to be throttled. For selected URLs, a given percentage of the first requests received after start-up are automatically throttled. If the request is throttled, a message is sent to the user that the server is busy but to try later. Servlets that are not throttled are allowed to flow through to be processed normally.

Probabalistic throttling of application server requests

An improved solution for intelligent configuration for restarting failed application server instances is provided. In an embodiment of the invention, a method for restarting a failed application server instance includes: receiving a notice of a failure of an application server instance; obtaining a cause of the failure; automatically applying at least one configuration change to the application server instance based on the cause; and recovering the application server instance.

Intelligent configuration for restarting failed application server instances

Embodiments of the present invention address deficiencies of the art in respect to portal view rendering and provide a method, system and computer program product for managing portal page response times. In one embodiment, a portal data processing system can include a portlet aggregator coupled to a portal server and configured to aggregate portlets in a portal page and to serve the portal page to a requester over a computer communications network. The system also can include a response time monitor coupled to the portal server and configured to measure response times both for the portal page and the portlets in the portal page. Finally, the system can include a response time remediation processor coupled to the response time monitor and configured to disable underperforming ones of the portlets in the portal page so as to improve the response time for the portal page.

Managing portal response times

A method and system for maintaining revision control of a set of objects within a data processing system, wherein the set of objects are linearly interrelated, such that a subsequent object is related to a preceding object. Each object is created, accessed and maintained within a hierarchial promotion structure having a number of levels. A first database is created, which includes data identifying the linear interrelation between objects within the set of objects. The data includes a revision identifier associated with each set of objects. A status indicator is associated with each object within the hierarchial promotion structure, wherein a user may access the set of objects utilizing a revision identifier or hierarchial promotion structure indicator.

Method and apparatus for maintaining revision contol of a set of objects

Library management systems are used in data processing systems as a repository for objects such as computer programming code, textual information and various other types of data. Any object may be taken from the library, modified and returned as a new version. Several new versions may be created, each of which may also have new versions created, making it difficult to obtain the 'correct' version. To resolve this difficulty, all versions are associated with their changes in order to determine whether a requested action, such as promoting a version from test to production, is allowable and truly desired. Thus, when an action is requested in terms of a number of reasons and a baseline version, any additional reasons for change associated with versions between the baseline and the desired version are determined. It is also determined whether there is a single version that encompasses exactly those versions created for the specified reasons from the baseline version.

Growing community of inventors

Raleigh, NC, United States of America

Stacy Renee Joines