Table of Contents
Advertisement
8-14.
FAULT LOG
The fault log contains events and faults that have
occurred during operation of the drive. These two
types of entries cover all situations except data er-
rors, which are recorded in the ERT log and run
time log, as previously described. The fault log
physically resides on the 00 maintenance track on
data surface O.
The distinction between a fault and an event is re-
lated to their impact on command execution. An
occurrence from which the drive can recover and
still complete the current transaction is classified
as an event. Events are retried by the drive in an
attempt to continue the transaction.
An occur-
rence that causes termination of a transaction is
classified as a fault. A fault represents a situation
from which the drive cannot recover, consequently
the failed operation is not retried.
During routine operation of the drive, all faults
are logged in the fault log. In addition, any event(s)
associated with each fault are also logged. Events
that are retried successfully and do not lead to a
fault are not logged in the fault log. For example,
if an off track condition (an event) is detected, the
drive will try to reposition the heads back on
track. If the attempt is successful, the off track
event is not logged in the fault log. However, if the
drive tries repeatedly without success to properly
position the heads, the operation is terminated and
a servo fault is logged in the fault log. All off track
events associated with the fault are also logged.
During an ERT, all events and faults that occur
are logged in the fault log. An event does not
have to be associated with a fault to be logged
during an ERT.
Faults and any related events are initially logged in
controller RAM, which holds a maximum of 30
entries. Periodically, the drive transfers the log
entries from RAM to the fault log on the disc
maintenance tracks. The disc fault log can hold a
maximum of 44 log entries. The drive returns both
RAM and disc entries when the fault log is read;
therefore, a total of 74 fault log entries can be
returned.
Each entry in the log contains the following
information:
Troubleshooting and Diagnostics
7936 and 7937
• Current address (cylinder, head, and sector).
This is the last address at which the servo sys-
tem was successfully positioned.
• Target address (cylinder, head, and sector).
This is the address the servo system was at-
tempting to access when the error occurred.
• Hardware Fault Register. This byte is a col-
lection of bits from the servo register and the
read/write controller register. The purpose
of this register is to provide additional in-
formation about the servo and read/write
systems when a fault in the data path occurs.
The significance of the individual register
bits is defined in table 8-1.
• Error Code. This is the TERROR/DERROR
(see table 8-2).
• Activity Number. This is a number which
represents the number of seeks within a range
that occurred between faults. The ranges are
listed in table 8-1. The numbers given are
approximations of the time that would have
elapsed between faults if the drive was
placed on a "typical" system and under "nor-
mal" activity.
8-15.
ERROR CODES
The drive encodes internal error conditions into
two-digit error codes. These codes, which are log-
ged in the drive status and fault log. are retrieved
using the appropriate external exerciser commands.
A complete list of the error codes is contained in
table 8-2. The table includes the error code, the
applicable controller interface, and a brief descrip-
tion of the error. The description identifies the er-
rors as Faults, Events, or Status Only errors. This
last category includes those errors which affect the
status of the current transaction, but are not
recorded in any logs.
Ta ble 8-2 also includes a list of assemblies most
likely to cause the associated error. The suspect as-
semblies are
listed
in
order
of
descending
probability.
8-5
Advertisement
Table of Contents
