ATT System 25 Maintenance Manual

							MAINTENANCE STRATEGY
Figure 5-1. Response to System 25 Trouble Report
5-2 
						

							MAINTENANCE STRATEGY
General Procedures
Sometimes, faults recorded in the error log and user-reported troubles will
exist at the same time. The logged faults should always be cleared first, if
possible. This procedure often clears the user-reported troubles without any
additional maintenance.
Clearing the most serious logged fault may clear some or all of the other
faults. Descriptions of the errors in the error log indicate their relative
urgency.
Error log records should be manually recorded before clearing a recorded
error or alarm. If a faulty CP is located in the process of clearing a problem, it
must be replaced with a known good CP. The associated error record should
be included with the CP when it is returned for repair. If the System
Administration Terminal (SAT) has an associated printer, the error records
can be printed instead of being manually recorded.
Traditional trouble-shooting methods still have a valid place in maintaining a
system as advanced as System 25. Indeed, they are sometimes sufficient to
locate and clear faults. These methods include voice terminal substitution,
visual inspections, continuity checks, and clarification of operating
procedures with users.
Total System Failures
As the flow chart in Figure 5-1 shows, a failure of the common control
circuitry (that is, Call Processor, Memory, and Service Circuit) must be
repaired immediately. This may clear other problems in the system as well.
Port Problems
Port CP failures can usually be cleared by replacing the CP displaying the
lighted red light-emitting diode (LED) (if reseating was unsuccessful). When
replacing suspected defective CPs, it is desirable to wait until the amber LED
(indicating CP is in use) turns off before removing the pack. (See
Interpreting Circuit Pack LEDs in Section 8, for additional information on
CP LEDs.)
5-3 
						

							MAINTENANCE STRATEGY
Common Control Problems
Common Control circuitry is difficult to troubleshoot and replace. The system
must be powered down before any of these CPs are replaced. Before
powering down, translations should be saved to tape (see Section 7
Operating the Digital Tape Unit). Other repair procedures may also require
the system to be powered down. After removing a suspected faulty CP, do
the following
l
l
lVisually inspect the CP. If a problem is identified, attach a note to the
CP identifying the problem.
Attach any error reports that may apply.
Visually inspect cabinet backplane.
After the suspected faulty pack has been replaced, the system can be
powered up. Replacing the Call Processor CP causes all historical alarm data
to be lost. The system will then have to be restored from the tape backup
unit with the latest copy of the systems translations. If no tape exists, the
system has to be reinitialized.
Also increasing the common control trouble-clearing difficulty is the high
degree of interaction between the CPs. Multiple errors and off-board errors
are more likely to occur in common control problems than in port problems.
Use of the error log is imperative in clearing common control problems.
Station, Wiring, and Trunk Problems
If the system indicates that a fault is with station, or if a user complaint
indicates a station problem, the trouble must be isolated to the station itself
or to the wiring between the cabinet or station interconnect panel and the
station.
A visual inspection of station wiring, particularly the mounting cord, is
suggested in cases of station trouble reports. A dead station can result from
the mounting cord being pulled loose. Station trouble can also occur when
wiring is crushed or severed by furniture or traffic. Stations can be swapped
with known good stations of the same type to help isolate the trouble.
Trunk problems reported by alarms or by users can be evaluated with tests
from stations or by interpreting the error log. If trunk troubles appear to be
outside the System 25, the responsible common carrier should be asked to
make repairs.
5-4 
						

							MAINTENANCE STRATEGY
Automatic Maintenance Tests
Because System 25 maintenance tests are performed automatically, there is
no provision for users or technicians to initiate maintenance tests. The error
log is read by entering commands from the SAT. Additional information that
may be of help in clearing complex or subtle troubles is also available via the
SAT (for example, the record of port locations).
Refer to Administration Manual (555-520-500) for your system for information
on using Search procedures from the SAT as an aid in troubleshooting.
Maintenance Failure
5-5
If a System 25 equipment problem cannot be corrected using the procedures
in this manual, the technician should follow established maintenance
escalation procedures. 
						

							ERROR LOG
The error log is accessible via the System Administration Terminal (SAT)
using the following procedures. The Error Log is comprised of the following
three error tables:
lPermanent System Alarms
lTransient System Errors
lMost Recent System Errors.
Each error record is reported as one line on the SAT (80 characters or
fewer). Permanent System Alarms and Transient System Errors tables use
the same format. These error records provide the location of the error [by
port and circuit pack (CP) type], the date and time of the first occurrence and
last occurrence of the error, the number of times the error has occurred, and
a description of the error.
A typical error record from the Permanent System Alarms and Transient
System Errors table is as follows:
PERMANENT SYSTEM ALARMS
PORT
BOARDFIRSTLASTCOUNTNAME
NUMBERCODEOCCURREDOCCURRED
10201ZTN82
23/12:3030/01:56
6
System Restart
Interpret this record as follows:
l
l
lPort Number 10201 is the 5-digit port identification number specifying
the location of the error. The first digit indicates the cabinet (1-3), the
next two digits indicate the slot within the cabinet (01-12), and the last
two digits indicate the port number on the CP in the slot (01-16).
Board Code ZTN82 is the CP on which the error occurred (or that is
associated with the error if this is an off-board error).
First Occurred 23/12:30 is the date and time of the first occurrence of
this error. The date is assumed to be the current year and month. A
24-hour clock is used.
6-1 
						

							ERROR LOG
lLast Occurred 30/01:56 is the date and time of the last occurrence of
this error. The date is assumed to be the current year and month. A
24-hour clock is used.
lCount 6 is the number of times this particular error has occurred.
lName System Restart is a description of the nature of the error.
Descriptions are self-explanatory.
Error records in the Most Recent System Errors table have a slightly different
format. A typical error record from the Most Recent System Errors table is
as follows:
MOST RECENT SYSTEM ERRORS
PORTBOARDDATETIMENAME
NUMBERCODEOCCURREDOCCURRED
10201ZTN82
03/12/8512:01:56ArchAngel Insane Restart
Interpret this record as follows:
lPort Number 10201 is the 5-digit port identification number.
lBoard Code ZTN82 is the CP on which the error occurred (or that is
associated with the error if this is an off-board error).
lDate Occurred 03/12/85 is the date this error occurred. A
conventional calendar designation is used.
lTime Occurred 12:01:56 is the time this error occurred. A 24-hour
clock is used.
A combined total of 40* (V1) or 50* (V2) error records for Permanent System
Alarms and Transient System Errors is stored by the system. Ten error
records are stored in the Most Recent System Errors table. Errors are
displayed in chronological order.
*
Two or more instances of the same error result in only two entries in the log (that is, the first
and last occurrence).
6-2 
						

							ERROR LOG
The error log is under the control of internal background maintenance
software that automatically escalates an error to a more serious category or
retires it. The system tests itself after a repair has been made. If the error
condition is corrected, the error may be removed from the error log.
Although some errors are automatically retired from the error log, they can
also be manually retired by entering commands at the SAT. Alarms,
conditions that light the alarm light-emitting diode (LED) on the attendant
console, can also be manually retired in the same way.
Accessing the Error Log From the SAT
Follow these steps to read any of the tables in the error log from the SAT:
1.
2.
3.
4.
5.
6.
7.Confirm that the SAT is properly connected to the Call Processor CP
administration port.
If the SAT has a selectable baud rate, set it to 1200 baud.
Be sure transmission parity is set to none, with the parity bit set to
space (0).
Note: For the AT&T Model 703 SAT, four pencil switches under
the paper compartment cover are set as follows: switches 1
and 3 to Off; switches 2 and 4 to On.
Turn on the terminal and press the carriage return character once or
twice. On most terminals, this key is  or .
After communication with the system is established, a prompt is
displayed:
Enter Password ->
Obtain the password from the system administrator. Enter the
password and press . The main menu is displayed with
the prompt: Make one selection from menu ->.
Enter 8 (for the SEARCH selection) and press .
6-3 
						

							ERROR LOG
8.The SAT displays:
When search is defined type c to search >Action=
9.Enter 1 and .
10.
11.
12.The SAT displays:
SEARCHES: ACTION=1 DATA=
Enter D. The system completes the word Data followed by the =
sign.
After Data= enter one of the following numbers, depending on which
error table you wish to see:
l 30 to see Permanent System Alarms
l 31 to see Transient System Errors
l 32 to see Most Recent System Errors.
If you select 30, for example, the system responds with: SEARCHES:
Action=1 Data=30.
Now, enter the letter c to begin the search of this table. The c must be
entered repeatedly to access each line.
To remove a record from the Permanent System Alarms or Transient System
Errors table, enter the letter r.
If you enter a letter incorrectly, press the backspace key. Each time you
press this key, one character is erased.
When you finish viewing one of the error log tables and want to view another
error log table, enter D. The system completes the word Data followed by
the = sign. Now enter the appropriate number for the table you want to view.
(See preceding Steps 11 and 12.)
If you want to return to the main menu at any point, type M and .
To sign off, simply turn off the terminal.
6-4 
						

							ERROR LOG
Error Log Dictionary
The following provides a listing of error messages displayed on the SAT and
an associated description of each.
AC Power Failure: An indication that ac power has been lost (for example,
plug removed from a wall receptacle, building power down, etc.)
Alarm Log Overflow: The error logs are full and no more errors can be
written to them. This error is kept, however, to alert you of the fact.
Arch Angel Insane Restart: Clocks may be missing from the TDM bus. Make
sure TDM bus terminators are plugged in. Insure that the Service Circuit CP
is plugged in and that the cables connecting the cabinets are okay. Check
Call Processor CP.
Arch Angel Interrupt Restart: System restarted due to Call Processor
problem. Check Call Processor CP if the error occurs frequently. Problem
could be software related.
Bad Data: (Other than CPU or Memory) Data portion on a down-link
message invalid. System may recover from this error.
Bad Major Heading: (Other than CPU or Memory) Header portion of a
down-link message is invalid. System may recover from this error.
Bad Oat Device: (STARLAN Interface CP) This error is an isolated transient
condition that is normally recoverable.
Bad Port: (STARLAN Interface CP) This error is an isolated transient
condition that is normally recoverable.
Bad Port Number: Port number on a down-link message is invalid. System
may recover from this error.
6-5 
						

							ERROR LOG
Bad Port Translations: Service Circuit or Tone Detector. Receiver requested
to listen to time slots that do not have tones on them. System may recover
from this error.
Bad Ringing Supply: Problems detected with ringing portion of power
supply.
Bad Sub Qualifier: (Other than CPU or Memory) Invalid message type on a
down-link message. System may recover from this.
Belated External Release: See No External Release On PBX Disconnect.
Board Initialization Audit Error: (Other than CPU or Memory) System may
recover from this.
Buffer Deallocation Failure: (STARLAN Interface CP) This error resets the
CP. If the reset is successful, then the error is transient and recoverable. If
the reset fails, the CP should be checked.
Bus Error Test Failed: Bus error circuitry failure. Replace CPU.
Cold Start: System was restarted for some reason (CP unplugged, power
problem, or some other error). Translations may be corrupted if the Call
Processor or Memory CP is removed while in use or if power supply shuts
down for any reason other than ac power failure (+5 V shorted, overheat).
The system initializes itself to the default state. A translation restoration is
required to reinstate customer specific data.
CPU ROM Checksum Error: CPU has failed the ROM test. Check Call
Processor CP.
DC Fail Check: Indicates the absence of +5 V dc on a cabinet. Each cabinet
has its own power supply; therefore, this may not cause a Warm Start.
DC Fail Transient Record: This message is provided when dc power fails. It
provides a record of the length of the outage.
6-6