Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Queue depth algorithm used for snapshot state monitoring

IP.com Disclosure Number: IPCOM000032468D
Original Publication Date: 2004-Nov-05
Included in the Prior Art Database: 2004-Nov-05
Document File: 1 page(s) / 44K

Publishing Venue

IBM

Abstract

Alerting of messages being in queues for too long can be hard, and can cause false alerts - or give alerts too late. This algorithm prevents false alerts and still enables timely alerts for even a small amount of messages on a queue.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 90% of the total text.

Page 1 of 1

Queue depth algorithm used for snapshot state monitoring

Disclosed is an algorithm that has been developed so only relevant alerts are being issued.

     Basically a snapshot of queue depth is being used and current depth is compared to the previous snapshot information.

     The algorithm uses 3 parameters: queue depth limit, (optional) delay, (optional) expected decrease.

The queue depth limit can be set to any level - even 0.

     If no wait time has been specified, an alert is given as soon as the queue depth limit is surpassed.

     If a wait time is specified, the alert is only given when the queue depth limit has been surpassed after the specified delay - except if an expected decrease has been specified.

     If an expected decrease has been specified, then no alert is given before the delay has passed and the decrease in queue depth is less than the expected decrease amount specified.

The full concept is illustrated below.

     Initially the depth is above the limit, second snapshot is taken within the delay time frame, so no alert is given.

     Third snapshot is still above the queue depth limit but the decrease is more than the expected decrease, whereas the fourth snapshot does not reach the expected decrease and thereby results in an alert.

     Warning messages can be used afterwards to track the progress and fine tune the parameters.

     Because the state is known, an information can also be given when the situation is back to normal.

queue depth

USER123W

USER 122W

Note that the specificati...