Browse Prior Art Database

Enhanced Method for Monitoring Critical Resources in Token Ring Networks

IP.com Disclosure Number: IPCOM000118381D
Original Publication Date: 1997-Jan-01
Included in the Prior Art Database: 2005-Apr-01

Publishing Venue

IBM

Related People

Prorock, TJ: AUTHOR

Abstract

This disclosure presents an enhanced method for monitoring Token Ring critical resources where: o Error rates are calculated for each critical resource. o Priority polling queues are established by the user to poll the highly reliable, medium, and low reliable devices. o Dynamic sorting of the critical resources into one of the priority polling queues based on actual device failure rates and user defined thresholds for each of the priority queues. o User configurable critical resource polling intervals. The user can determine the amount of time spent on polling critical resources in each of the priority polling queues. o Tracks failure rates for each monitored critical resource.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 14% of the total text.

Enhanced Method for Monitoring Critical Resources in Token Ring Networks

      This disclosure presents an enhanced method for monitoring
Token Ring critical resources where:
  o  Error rates are calculated for each critical resource.
  o  Priority polling queues are established by the user to poll
      the highly reliable, medium, and low reliable devices.
  o  Dynamic sorting of the critical resources into one of the
      priority polling queues based on actual device failure
      rates and user defined thresholds for each of the priority
      queues.
  o  User configurable critical resource polling intervals.  The
      user can determine the amount of time spent on polling
      critical resources in each of the priority polling queues.
  o  Tracks failure rates for each monitored critical resource.
      Provides the capability to perform trend analysis based on
      the historical failure rate statistics.

      The main idea is that, over time, all devices will adaptively
be placed in a queue in accordance with their determined behavior.
The result is that those critical devices that tend to be faulty will
be monitored more closely than those devices that prove themselves to
be highly reliable.  The disclosed algorithm adjusts/adapts
dynamically to  ensure the network trouble-spots are monitored at all
times.

      There are several IBM* program products to manage Token Ring
Segments.  The two premier products are LAN Network Manager for
OS2(LNM/OS2) and LAN Network Manager for AIX(LNM/AIX).  Both of these
products provide a function called critical resource monitoring.

      Critical resource monitoring operates as follows:  from a list
of ring stations on a token ring segment, customers can designate
specific stations as critical resources.  Typically, these resources
are bridges, routers, or file servers that are vital to an
enterprise's mission.  Then, the critical resource monitoring is
accomplished by polling a critical resource at periodic poll
intervals to ensure the resource is online and available.  LNM/OS2
allows for 1000 critical resources to be defined.  The critical
resources are processed (polled)  in a "round robin" fashion through
the entire list and the result of the  poll is a simple go/no-go
indication that the Critical Resource is still  operational.

      The problem is that although the existing Critical Resource
processes provide the necessary function, no special consideration is
given to any of the resources.  They are all treated as equal and
processed sequentially every poll period, and the results of the poll
are a simple pass/fail notification.

      The problem is that in reality some critical resources are more
error prone than others.  Some of the newer network devices are very
reliable with low specified Mean Time Between Failure (MTBF) values.
Also, the reliability is increased even further in some of th...