Retry method and apparatus for use in a magnetic recording and reproducing system
First Claim
1. A method for retrying commands received by a magnetic tape peripheral controller from a data processing unit coupled to said controller during the handling of error conditions occurring during the execution of read and write operations respectively in which information bytes of digital signals of a block having a predetermined format are read from a multichannel magnetic medium and written by a magnetic tape device on said multichannel magnetic medium, said device coupled to said controller, said controller including processing means, recovery means coupled to said device for assembling said digital signals into said bytes, said recovery means including activity detector means responsive to said signals to generate signals indicative of the rate at which said signals are being received, and indicator storage means coupled to said detector means and to said processing means, said method comprising the steps of:
- a. generating output signals by said activity detector means during the monitoring of the reception of said digital signals received by said recovery means from said tape device corresponding to said bytes of said block being read by said tape device from said medium for a predetermined number of time intervals during the execution of an operation by said controller processing means specified by a command received from said data processing unit;
b. storing a plurality of error signal indications in said indicator storage means in response to said output signals from said activity detector means which define the profile characteristics of said block detected during step (a); and
,c. generating signals by said controller processing means for initiating a retry of said command and re-execution of said operation upon the completion of said operation upon detecting an error condition only when said error signal indications stored in indicator storage means signal that said profile characteristics of said block read during said operation indicate that said device can reliably retry said operation.
0 Assignments
0 Petitions
Accused Products
Abstract
A magnetic tape subsystem includes a peripheral controller which processes data signals of a block received during the reading and recording of the block on a magnetic tape medium by a selected one of a plurality of magnetic tape devices in response to commands received from a data processing system. The peripheral controller includes a data recovery unit having a plurality of storage indicators. These indicators are set by the controller in accordance with the characteristics of the block of data signals recovered by the data recovery unit providing indications of the results of the reading or recording operation performed by the selected magnetic tape device during the peripheral controller'"'"'s execution of the command. The controller divides up the interval of time of an operation into a number of time periods and monitors the characteristics of the signals of a block which should be received by the data recovery unit during those intervals. During execution of a write command, the controller monitors the tape recording or write operation during different time intervals by monitoring the continuity of the signals recorded by the magnetic tape device indications of which the data recovery unit receives during its performance of a read after write checking operation. The various indicators signal whether the block written was normal or was damaged in addition to signaling the extent of damage which occurred during the writing of the block. Signals representative of the indications are grouped to specify at least two types of error conditions which are tested by the controller to establish whether the command is retryable and nonretryable. The peripheral controller initiates a retry of the command executed only when the indicator signal that the extent of damage has not exceeded a predetermined amount thereby ensuring that only commands which can be retried successfully are re-executed. The controller performs similar monitoring operations during the execution of read commands for establishing signal indications used in determining whether the command should be retried upon detection of errors in the block read by the data recovery unit.
-
Citations
39 Claims
-
1. A method for retrying commands received by a magnetic tape peripheral controller from a data processing unit coupled to said controller during the handling of error conditions occurring during the execution of read and write operations respectively in which information bytes of digital signals of a block having a predetermined format are read from a multichannel magnetic medium and written by a magnetic tape device on said multichannel magnetic medium, said device coupled to said controller, said controller including processing means, recovery means coupled to said device for assembling said digital signals into said bytes, said recovery means including activity detector means responsive to said signals to generate signals indicative of the rate at which said signals are being received, and indicator storage means coupled to said detector means and to said processing means, said method comprising the steps of:
-
a. generating output signals by said activity detector means during the monitoring of the reception of said digital signals received by said recovery means from said tape device corresponding to said bytes of said block being read by said tape device from said medium for a predetermined number of time intervals during the execution of an operation by said controller processing means specified by a command received from said data processing unit; b. storing a plurality of error signal indications in said indicator storage means in response to said output signals from said activity detector means which define the profile characteristics of said block detected during step (a); and
,c. generating signals by said controller processing means for initiating a retry of said command and re-execution of said operation upon the completion of said operation upon detecting an error condition only when said error signal indications stored in indicator storage means signal that said profile characteristics of said block read during said operation indicate that said device can reliably retry said operation. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20)
-
-
21. A method for retrying commands executed by a peripheral controller received from a data processing unit involving any one of a plurality of storage devices coupled to said controller, each of said storage devices having a storage medium on which blocks of digital signals are recorded, said commands being retried in response to error conditions occurring during input/output operations specified by said commands in which digital signals of a block are read from said storage medium or written on the storage medium of a selected one of said plurality of storage devices, said controller including processing means, data recovery means coupled to said plurality of said devices and storage means coupled to said recovery means and to said processing means, said method comprising the steps of:
-
a. reading from said storage medium, information bytes of said block during the execution of an input/output operation by a selected one of said devices specified by one of said commands received from said unit; b. transferring signals representative of said bytes by said one device to said controller recovery means; c. storing by said processing means a plurality of signal indications in said storage means in response to said signals received by said recovery means denoting the profile characteristics of said block being read from said medium during said execution of said operation by said device; d. grouping by said processing means different ones of said signal indications in said storage means into to at least first and second categories of conditions; and
,e. initiating the start of a retry of said input/output operation by said processing means upon completing execution of said operation only upon detecting that none of the indications in said first category has been stored in said storage means signaling that said operation can be retried reliably by said device. - View Dependent Claims (22, 23)
-
-
24. A data processing system including an input/output processing unit including storage for a plurality of channel command entries (CCE'"'"'s), each entry including a plurality of control bytes, one of said bytes corresponding to a command byte coded to specify an operation to be performed, a microprogrammed peripheral controller coupled to said input/output processing unit and responsive to command control bytes from said processing unit to perform input/output operations involving a selected one of a plurality of storage devices coupled to said peripheral controller, said controller comprising:
-
an addressable control store for storing microprograms including a plurality of microinstructions used for transferring bytes of a block between said input/output unit and a selected one of said devices in response to one of said command control bytes; decoder means coupled to said control store for generating control signals in response to decoding microinstructions read out from said control store; data transfer means operatively coupled to receive signals from said selected one of said devices during the execution of said operation to be performed specified by one of said command bytes, said data transfer means including activity detector means coupled to receive said signals and operative to generate output signals indicative of the rate at which said signals are being received; and
,signal indicator storage means coupled to said detector storage means and to said decoder means, said signal indicator storage means being conditioned by signals from said decoder means during a predetermined number of time intervals occurring during execution of said operation to store a plurality of signal indications which define the profile characteristics of said block being read from said device as signaled by said activity detector means output signals; and
,said controller including means operative upon detecting an error to address said control store for read out of microinstructions for generating signals to initiate retry of said command by said device only when said signal indications stored in said indicator storage means representative of said profile block characteristics signal said controller that retry of said command can be executed reliably. - View Dependent Claims (25, 26, 27, 28, 29, 30, 31, 32, 33, 34)
-
-
35. A peripheral controller coupled to a data processing system and to a plurality of magnetic tape devices connected to read and write frames of digital signals arranged in blocks on a number of channels of a magnetic medium, each block having a predetermined format, said controller including processing means, an addressable control store for storing microprograms used to perform read and write operations in response to commands received from said processing system, decoder circuits coupled to said control store for generating control signals in response to decoding microinstructions read out from said control store, a data recovery unit for assembling into bytes digital signals transferred by a selected device, said recovery unit having an input coupled to receive said digital signals from said selected device and an output coupled to said processing means and including a plurality of error indicator storage circuits operative to store signals indicative of error conditions occurring during the assembling of said digital signals into bytes and activity detector means coupled to said input and operative to generate signals indicative of the rate at which said digital signals are being received, said controller further including retry apparatus comprising:
-
signal indicator storage means coupled to said activity detector means and to said processing means; said control store further including a plurality of microinstructions of a retry routine and a plurality of predetermined sequences of microinstructions included in said microprograms for monitoring signals generated by said activity detector means; said decoder circuits being operative during the execution of said operation to generate signals in response to read out of said predetermined sequences of microinstructions for conditioning said signal indicator storage means during a predetermined number of time intervals during execution of an operation to store signal indications which define a profile of said block being transferred in accordance with said signals generated by said activity detector means; and
,said controller processing means upon detecting an error condition stored by said error indicator storage circuits to address said control store for read out said microinstructions of said retry routine to initiate retry of said command only when said signal indications stored by said indicator storage means representative of said profile signals indicate to said controller processing means that said selected device can retry said command reliably. - View Dependent Claims (36, 37, 38, 39)
-
Specification