AutoBER Program and Refcode
The automatic BER analysis program (autoBER) is a facility for problem source
identification. Each BER is analyzed by the autoBER program to produce an
eight-character reference code (refcode). The latter is included in the alarm and
alert messages. This refcode helps identify a failing hardware or software
component, or some user error (NCP generation, CDF upgrade, and so on).
When the problem determination action provided for the alert is not sufficient to
isolate the problem, the user should provide the refcode to the IBM maintenance
organization for failure analysis and isolation.
Generic Alerts
Alert messages contain generic information formatted by the NetView program.
They contain a recommendation for further problem determination actions (tests or
calling the appropriate service organization).
Alert messages are divided into two parts, a coded part with the refcode that
contains information about the failure, and a text part that contains a general
description of the failure. The coded part is translated by the NetView program in
the host into a message displayable at the host console. The refcode corresponds
to a precise type of error and gives information about the suspected component.
When the NetView program is not supported or not active, the error is reported as
a record to the VTAM log file (Logrec) and an alarm message is displayed at the
3745 operator console.
Alarms
Alarm messages displayable at the 3745 console provide the user with an entry
point into the IBM 3745 Problem Determination Guide, SA33-0096. It is
recommended that the 3745 operator review the alarms as a chronology of events
before taking action on any specific alarm. MOSS records all alarms on the disk.
Retry and Recovery
Error handling in the 3745 includes extensive recovery and retry for intermittent
failures. On detection of a failure, the information on the error environment is
logged and a retry or recovery attempt is invoked.
Retries allow the controller to recover from intermittent hardware and transmission
failures. In most cases, the controller remains available to the rest of the network
while the retries are being executed. The controller provides recovery from
intermittent hardware failures. The host access method provides network error
recovery.
The failed operation is retried the number of times defined by the function or
component threshold value.
According to the level of error, retries are performed by:
NCP
Scanner microcode
MOSS microcode.
5-4
IBM 3745 Models 130, 150, 170: Introduction
Summary of Contents for ThinkPad 130
Page 1: ...IBM 3745 Communication Controller Models 130 150 and 170 IBM Introduction GA33 0138 2...
Page 2: ......
Page 3: ...IBM 3745 Communication Controller Models 130 150 and 170 IBM Introduction GA33 0138 2...
Page 8: ...vi IBM 3745 Models 130 150 170 Introduction...
Page 15: ...IBM 3745 Communication Controllers in a Typical Network Chapter 1 Highlights 1 3...
Page 24: ...2 4 IBM 3745 Models 130 150 170 Introduction...
Page 32: ...3 8 IBM 3745 Models 130 150 170 Introduction...
Page 72: ...5 8 IBM 3745 Models 130 150 170 Introduction...
Page 74: ...6 2 IBM 3745 Models 130 150 170 Introduction...
Page 78: ...3745 130 150 170 General Data Flow 6 6 IBM 3745 Models 130 150 170 Introduction...
Page 82: ...A 4 IBM 3745 Models 130 150 170 Introduction...
Page 86: ...X 4 IBM 3745 Models 130 150 170 Introduction...
Page 92: ...X 10 IBM 3745 Models 130 150 170 Introduction...
Page 94: ...X 12 IBM 3745 Models 130 150 170 Introduction...
Page 100: ...X 18 IBM 3745 Models 130 150 170 Introduction...
Page 103: ......