Error control in a hierarchical system
AssigneeInternational Business Machines Corporation
InventorsBrent C. Beardsley, Allen C. Brailey
Life of this patent
Estimating the term…
Abstract
A hierarchical communication system has multipaths for different levels of the heirarchy, each set of paths is assigned a criticalness to the successful operation of the system. Error rates for all of the paths are monitored. A threshold for defining an unusable data path is based upon the criticalness of the path to successful operation. That is, the more critical the path, the higher the error rate that will be sustained. A specific embodiment employs shift registers for indicating the error rate of the last predetermined number of usages of the given paths. A mass storage system employing the error-rate system is described.
Claims10
The method of operating a multilevel communication network having first and second levels each with a plurality of paths, a plurality of units selectively interconnecting said levels such that one path in each level is coupled via a unit to create a signal transfer path through said levels and a one of said units,
- the machine-executed steps of:
- assigning a criticalness for successful network operation to each path in both said levels, said criticalness being different in such levels;
- measuring the error rate of each path over a predetermined number of path usages;
- setting a threshold of error-rate acceptability for said paths in accordance with said assigned criticalness, the greater the criticalness, the greater the threshold;
- comparing said measured error rates with said thresholds, respectively, for said paths;
- and
- when said measured error rate exceeds said threshold for a given path, reconfiguring said network to eliminate said path therefrom.
In a multilevel peripheral system having a first set of director units adapted to be attached to hosts,
- a second set of device controls,
- a first set of signal paths extending between said director units and said device controls such that multipathed connections between said director units and said device controls is enabled,
- a third set of peripheral devices,
- a second set of signal paths extending between said device controls and said peripheral devices such that multipathed connections between said device controls is enabled,
- control means being coupled to all of said device controls, said devices and said paths for enabling accessing said paths to establish said multipath enabled connections, the improvement including in combination:
- a plurality of digital shift register means, each said digital shift register means having a first predetermined number of digit positions, one of which is a digit entrance and another is a digit exit and the digital shift register means shifting digits stored therein from said entrance toward said exit;
- a one of said digital shift register means being for accumulating error rates for each of said paths,
- error monitoring means for each of said paths in both said first and second sets and being responsive to predetermined errors in said paths to set an error-indicating digit in said digit entrance, respectively:
- plural access monitoring means operatively associated with said paths, respectively, to shift the digit contents of said digital shift register means, respectively, for each path connection;
- first and second error rate monitor means operatively associated with said digital shift register means for said paths in said first and second sets, respectively, and including first and second error rate thresholds, respectively, to indicate an unacceptable error rate for a given one of said paths;
- and reconfigure indicating means being coupled to both of said error rate monitor means for indicating said given one path is unusable.
A plural unit data processing system having a plurality of paths extending between said units in a manner that alternate access paths between diverse units in the system are available,
- the improvement comprising:
- means for indicating criticalness of each path with respect to successful system operation;
- a plurality of error rate means coupled to said indicating means and respectively operatively associated with each path and having an error rate threshold in accordance with said predetermined indicated criticalness of each path and means for indicating rate or errors exceeding said thresholds, respectively;
- means for monitoring errors on said paths and supplying electrical indications of such errors to the respective error rate means;
- and
- reconfiguration means being coupled to said monitoring means and to said units and paths for being responsive to said rates exceeding said threshold (RET) indications to reconfigure the system by deleting said path with the RET indication out of the system.
The machine-implemented method of controlling a hierarchical system having plural levels, one of the levels being an access level and another being a device level with intermediate levels therebetween;
- a plurality of access paths extending between said levels to provide access to said device level from said access level via said intermediate level on alternate ones of said access paths,
- the steps of:
- for each path, establishing a permitted error rate,
- repetitively attempting to access said device level;
- upon each attempted access to said device level, detect and log errors beginning at said access level such that error conditions closest to said access level are detected in any one access to the exclusion of error conditions at levels closer to said device level;
- and
- reconfiguring said system based upon said detected and logged error rates such that errors occurring at the device level are detected only when in a succession of said repetitive attempted accesses.
Dependencies are inferred from the claim text.