ST01: - SYSTEM TRACE
ST01 is the system trace, we can see the
1) Authorization check: -If this
switch is activated, all access authorizations that the system performs are
recorded. The resulting report shows when the system checked what authorization.
Double-click
on a trace record to get the following information:
- Date,
time
- Work
process number
- User
- Authorization object
- Result (0, if authorization was
given)
- Program, line
- Number of authorization values
The authorization values are then
listed.
2) Kernal Functions: -Trace Switch: Calls from C Modules
Calls of kernel function.
Some kernel functions write this
trace.
3) General
Kernel: - Trace switch: Unformatted entries
from C programs
4) DB
Access SQL Trace: - Trace switch:
"Actual" database accesses
This trace switch concerns the interface between the platform-independent parts of the
database access module and the platform-dependent parts.
Here you can see what RSDB could not achieve via SAP buffers.
5) Table
Buffer Trace : - Trace switch: Buffered tables
6) RFC
calls :- Trace switch: RFC calls
7) Lock
operations: - Trace switch: Enqueue calls
Depending upon the requirement we need to switch on the
trace and after specific time switch off the trace. If you keep on the trace,
the trace files may grow unlimitedly there by disk overflow may occur.
ST02: -BUFFER-TUNNING SUMMARY
ü Buffers
are need to monitor regularly, in maintaining buffers in the application
servers improves the performance of the system.
ü During
restart the system the performance will be low because the buffers are cleaned
during restart. That is reason the system perform need to monitor after the
buffer accumulation.
ü The
various types of buffers like Name Tab, Program Buffers, CUA, CalendarBuffers,
Table Buffers, and Screen Buffer each of the areas are assigned with certain
instance profiles.
ü There
is also a column called Swaps, which
is critical. If there is more number of SWAPS, you need to identify the reason.
ü The
reasons for Buffer Swaps are
o If
there is no memory
o if
there is no directories
ü Based
on the SWAP allowable limit you need to change the size of the buffers. While
changes the buffer sizes keep in ‘KIV’, buffers sizes should be allowable in
the total memory.
ü You
can identify all existing parameters in the current parameters. There is a
memory column which shows the amount of memories utilizes in the system.
Various
memories are
o Heap
memory
o Extended
memory
We have to see that system
doesn’t use complete heap memory and extended memory.
By clicking detailed analysis
we can see the following figure
-It
will display
Ø Also gives detail of SAP memory, roll area, paging area,
extend, heap memory
Ø Types of buffering:
Ø Full buffering
Ø Single record buffering
Ø Generic record buffering
Ø No buffering
Ø Improve buffer hit ratio (>94%)
Ø Check buffer utilization.
Ø Improve buffer utilization by increasing buffer size
(ABAP/buffer size) thro RZ10.
Ø Maximum allowable limit for swaps are recommended by SAP
is 5000 to 10000.if it is more than 20000 then increase buffer size.
STO3:- WORKLOAD ANALYSIS
ü The
load analysis of the system will be monitor in t-code ST03.
ü Go
to Expert mode- select day, weak or month load.
ü The
amount of time consumed by individual components like roll time, wait time, CPU
time, DB time … etc. will be displayed for each user, transaction, programs and
fro each processes. Average is also displayed for the above components.
ü Good
response time of a dialog step will be in between 600 m.sec to 1800 m.sec.
ü Roll-in
time and roll-out time should be more than 10% of the response time.
ü Load
time, CPU, DB times should be more than 40% of (RT – WT)
ü If
CPU time is more, look into ABAP programs
Response
time per dialog step:
The response time of a dialog step is the time required
for requesting the dialog from the dispatcher work process to the processing of
the dialog through the ending of the dialog in the dispatcher and the transfer
of the data to the Presentation Server.
This also includes the time used for
"Roundtrips" to transfer data from the SAP R/3 front-end to the application
server and back. This time is recorded as roll wait time on the application
server, while it is displayed as GUI time on the front-end. For inefficient
networks, such as WAN connections, this time can contribute significantly to
the response time, although it uses no resources on the application server, as
the context is rolled out and the work process is released.
The response time is usually split into wait time and
"dispatched" time. The SAP response time is made up of the following
components:
+ Load times for programs, screens, and GUI interfaces
+ Roll times for rolling in work data
+ ABAP processing times
+ Database time
+ Enqueue time for logical SAP locks
+ Roll wait time (not including task types RFC/CPIC/ALE).
The CPU time is not an additive
component of the response time, but rather the sum of the CPU time used by the
individual components. The CPU time is an independent additional piece of
response time information.
ST04 – DATABASE MONITORING
Here we can see Buffer busy waits, file system requests,
wait events, sap client, Oracle session, SQL request, table access, exclusive
lock waits, latch waits, Database message log, state on disk, parameter
changes, performance database, summary report
Db cache should be greater than 98%
User/recursive calls less than 5
Reads/user calls greater than 40
Buffer busy waits.
Ø D/b monitor.
Ø
Used to
monitor the performance of database.
Ø Database Buffer Hit Ratio should be >94%.
Ø When it is < than 94% the data is fetched directly from
the d/b. So this affects the performance of the system.
Ø Hit ratio = [(Logical read-physical read)/Logical
read]*100.
ST05: - SYSTEM PERFORMANCE
-> With this transaction you can switch different
traces on or off, output trace logs in a list, display further information
about the logged statement (such as an SQL statement),and create the execution
plan for an SQL statement.
-> You can also create, save, and study the performance
of SQL statements. For information on other functions, see the documentation
(see below). You can gain an overview of the available functions from the menu
entries.
-> By choosing Help -> Help on statement you can
display the documentation on the SAP R/3 System trace.
-> If you would like to receive messages during the
evaluation of the performance trace (for example, a progress display for the
trace record formatting), enter the command "MESSAGE" in the OK code
field. You can terminate the performance trace after 5000 records have been
formatted by entering "CHECK" in the OK code field. You can reset
these settings by entering "RESET" in the OK code field.
ST06 – OPERATING SYSTEM MONITORING
Here we can see
Ø -O/s monitoring.
§ Check CPU idle time memory, swap
§ Disk with highest response time.
§ Check CPU, Memory, Swap, DISK, LAN.
§ Used to comparing with previous days or previous hours.
§ CPU idle time >= 30%.
§ If it is low it indicates that load on the system is high.
§ Analyze and reduce the load.
§ Gives statistical information from the O/S.
§ Detailed analysis can be performed.
§
Restrict the
consuming of high CPU resources.
Ø Number of
CPUs in the system
Ø CPU load
average for the last 1 minute
Ø CPU load
average for the last 5 minutes
Ø CPU load
average for the last 15 minutes
Ø Physical
memory available
Ø Physical
memory free space
Ø Number of
pages, paged in and out per second.
Ø Maximum
swap size (by file system swap: freespace include)
Ø Actual size
of total swap in KB
Ø Response
time of the disk
Ø Disk with
highest response time
·
Name
·
Utilization
·
Avg wait time per m.s
·
Kb transferred per sec
Ø Here we can
start and stop saposcol
The operating system collector SAPOSCOL is a stand-alone
program that runs in the operating system background. It runs independently of
SAP instances exactly once per monitored host. SAPOSCOL collects data about
operating system resources, including:
Ø Usage of
virtual and physical memory
Ø CPU
utilization
Ø Utilization
of physical disks and file systems
Ø Resource
usage of running processes
ST07: - APPLICATION MONITORING
Ø Number of
users of an SAP application
Ø Number of
active users of an SAP application
Ø Number of
users connected to one work process
Ø average
number of sessions per user
Ø Number of
servers on which an SAP application is running
Ø Here we can
see all the above information for a particular client also.
By Click on
Userdistribution – change client
Ø And also
for particular application server.
By Click on Userdistribution –
choose.app.server
Ø Here we can
see particular module of sap buffer, DB access, DB memory
Ø We can view
Response time of previous days, previous week, this month, previous month.
Here we can see where the tables buffered in not buffered,
single key buffered, generic key buffered.
ST11- ERROR LOG FILES ( DEVELOPER TRACE )
Ø
Developer
traces contain technical information for use in the event of problems with your
system. Using the entries in the developer traces requires a sophisticated
knowledge of the host systems in which your SAP system is running and of the
SAP system itself.
Ø The
traces can be useful in diagnosing host system and SAP-internal problems that
are affecting your SAP system.
Ø Developer
traces are written in files in the work
directory of the SAP application server that generated the trace.
Dispatcher dev_disp
Workprocess dev_w0
to wn
Gateway dev_rd
R3trans and tp transport programs dev_tp
Monitoring
infrastructure (test mode only) dev_moni
Activating
/ Deactivating Developer Traces from within SAP System:
Enter transaction code SM50.
Choose the work process in which you wish to increase the
trace level.
To trace all work processes of a server, use the system
profile method shown below.
Choose Process ® Trace ® Active components.
The system presents a dialog screen that shows the current
status of the developer trace.
Turn developer tracing on and off for different server
components by selecting the appropriate checkboxes.
Set the degree of detail by entering a number in the Level
field. Possible trace levels are as follows:
– 0: no trace.
– 1: write error messages in the trace file.
– 2: full trace. The trace entries that are actually
written can vary with the SAP program that is being traced.
- 3: additionally, trace data blocks.
You can also set trace options instance-wide with the rdisp/TRACE=<n> option. The trace values are the
same as those in the list above.
ST22:- MONITORING ABAP DUMPS
The runtime errors of the system will be logged in dump
analysis. Whenever there is programmatically error [or] reports exceeds maximum
for time execution, table space overflow, max extents reached will be logged
into the screen.
(1)
What happened?
(2)
What can we do?
(3)
Error analysis?
(4)
How to correct the
error?
(5)
System
environment.
(6)
Information on
where terminated?
Analysis the
requirement to run that Program/report and run the program again.
Try to debug the problem with the help of error message
get assistance from seniors. Search SAP market place the problem still persists
log a message to SAP.
License with SAP and try to get the solution and implement
it. If necessary we will follow the landscape for applying the recommended
notes.
DB01: - EXCLUSIVE LOCK WAITS
Oracle: Waiting for
exclusive database locks (exclusive lock waits)
An exclusive database lock is set when a user locks a
table record. If a
second user tries to process this record, he or she must
wait until the
first user releases it. This situation is an 'exclusive
lock wait'. This
monitor shows such wait situations. You can display an
overview of all
database locks (V$Lock) by choosing 'V$Lock'.
The fields are:
Object : Name of the locked
table
for lock holder and lock waiter:
SID : Oracle session ID
SPID : Oracle shadow process ID (at operating
system level)
Client host: Name
of the host on which the R/3 work process is running
PID : R/3 work process ID (at operating system
level)
Level : Level > 1: Lock holder is waiting for
another lock
Time : Time
(in seconds) since a lock was set or wait time
Row ID: Oracle row
ID of the locked record
Level = 1 : Lock Situation detected
Level > 1 : Lock Holder is waiting on one or more Locks
itself
Level = -1: Deadlocks detected
DB02 :- DATABASE PERFORMANCE {TABLES AND INDEXES}
In this
transaction we can see
Database and Tablespaces
Ø Date and time of the last analysis
Ø Total number of tablespaces in the database
Ø Total size of all tablespaces in KB
Ø Free space in all tablespaces in KB
Ø Used space of all tablespaces as a percentage
Ø Free size of the tablespace with the lowest amount of free
space in KB
Ø Used space of the tablespace with the highest fill ratio
as a percentage
Ø Current size
Ø Space statistics
Ø Free space statistics
Ø D/b size should not above 90%
Tables and Indexes
Ø Date and
time of the last analysis
Ø Total
number of tables defined in the database
Ø Total
amount of used space of all tables defined in the database
Ø Total
amount of used space of all indexes defined in the database
Ø Total
amount of used space of all indexes defined in the database
Ø Date and
time of the oldest execution of RUNSTATS using
program dmdb6srp
Ø Date and
time of the latest execution of RUNSTATS using
program dmdb6srp
Ø Detailed
analysis on tables and indexes
Ø Missing
Indexes
Ø Space
critical objects
Ø Space statistics
DB12: - BACKUP LOGS
We can see in this transaction
DATABASE BACKUPS
Ø Last
unsuccessful backup ( time, date, and with return code)
Ø Last
successful backup
Ø Overview of
database backups
REDO LOG BACKUPS
Ø Archiving
directory statistics ( free space)
Ø Overview of
redo log files
Ø Overview of
redo log backups
Report: RSORA850
You can use this program to display BRBACKUP and
BRARCHIVE logs (Database backup and archive of redo log files).
DB13: - DBA PLANNING CALENDAR
Ø You can use the DBA Planning Calendar to automate Oracle
database administration. This includes implementing, executing, and checking
actions
Ø You can use the DBA Planning Calendar for almost all
regular database administration actions. This includes tasks for which the
Oracle database system must be stopped, such as offline backups.
Ø You can only use the DBA Planning Calendar to start
actions if the R/3 System is up and running. You can use the BRtools or SAPDBA
to execute tasks such as recovery, for which the SAP System must be inactive.
Ø All DBA
activities that stop the database, such as offline backups, will terminate
active SAP transactions. Schedule such activities for night runs, and warn
users of the interruption using utilities – system messages
Ø In all such tasks, the DBA Planning Calendar stops the
database and starts the action automatically. The SAP System is not available
as long as the database is stopped. However, the SAP System itself is not
stopped. Once the database is available again, the SAP System is automatically
reconnected.
DB14: - Display DBA Operation log for Database
DBA
SAPDBA(sapdba
operations)
Ø SAPDBA
Detail log for DATABASE
BRCONNCET(brconnect operations)
Ø Number of
Tables Without Statistics:
Ø Number of
Indexes Without Statistics:
Ø Number of
Tables Whose Statistics Were Deleted:
Ø Number of
Indexes Whose Statistics Were Deleted:
Ø Number of
Tables Whose Statistics Were Checked:
Ø Number of
Tables Selected After Check:
Ø Number of
Tables for Which Statistics Were Collected:
Ø Number of
Indexes for Which Statistics Were Collected:
Ø Number of
Indexes Whose Structure Was Checked:
BACKUPS
BRBACKUP(Database backups)
Ø BRBACKUP
action log of database instance
Ø BRBACKUP
detail log of database instance
BRARCHIVE(Redo log Backups)
Ø BRARCHIVE
action log of database instance
Ø BRARCHIVE
detail log of database instance
OTHERS
Ø Other
Operations
Ø Non-SAP
Data Archiving
Ø All
Operations
Ø All
Function IDs:
DB20: - EDIT TABLE STATISTICS
Here we can
Ø Create
statistics of table
Ø Delete
statistics of table
Ø Check
structure of table
Ø DBSTATC Maintenance
Ø DBA
Operations(transaction DB14)
Ø Start
Global Statistics Operation:
Update Step 1: Delete Harmful Statistics
Step 2: Check and Update
and Generate Initial Statistics.
All steps are executed in accordance with the
settings in table DBSTATC.
Update(DBSTATC) Check and Update Statistics
of DBSTATC objects.
If the TODO indicator is set, the statistics
for
the table are updated without a check.
Create missing statistics Creates
statistics for tables that should
have statistics, but do not.
Delete Harmful statistics Deletes
statistics for tables that should not
have statistics, but which do have some, such
as pool and cluster tables.
DB24:- Logs for Administrative Database Operations
Here we can do
Ø All complete logs or jobs.
Ø Displays optimizing statistics.
Ø
Update
statistics.
Ø Backup logs.
Ø Reorganization
Ø Consistency check
*SM51:-
Application overview
-It
will displays
|
Server name
|
Host name
|
Type
|
Status
|
-Status
may be
(a) Starting
: Application server are being started. They cannot process any request yet.
(b) Initial:
The server has logged on to Message server but cannot be addressed yet.
(c)
Active: Processing the request.
(d) Passive:
Application server will be deactivated; it will complete the processing of task but
will not accept any new tasks.
(e)
Shutdown: Server is being shutdown.
(f)
Stop: Server has no connection with the message server.
What we should do?
- Check the status
- Which client, user, Report and action on the user worked on work process
- Reasons for the status of work process hold, stopped,
running, wait; start(y or n)
- Check the no. of servers
- If it is stopped click on the release notes to find
Operating system, Database and Kernel Version.
- Double click on the server it will display SM50
*SM50:- Workprocess overview
-It will displays
|
S.no
|
Type of w/p
|
PID
|
status
|
Reason start
|
Err
|
CPU time
|
Report
|
Client
|
User
|
action
|
Table
|
W/p status
may be,
-Runnig,
Waiting, Holding, Terminated
W/p Type may
be
DIA, UPD, UPD2, ENQ, BTC, SPO
What we
should do?
§ Check Yes/No i.e. Restarted or not Restarted
§ Check the error how many times a W/p terminated abnormally
§ If all the W/p running stage identified the W/p which is
consuming more time (Monitor the work process at O/s level using Dpmon
§ Go task manager and the PID kill the process.
§ Identify work proces which is consuming more heap memory
in sm66 and inform the user before terminating workprocess. Every action will
be done through CRF(change request form) or notification to the user will be using
the t-codes
SM66 : - Global Active Workprocess Overview
-It will
displays
|
Server name
|
Type
|
PID
|
Status
|
Reason start
|
ERR
|
CPU time
|
Clnt
|
User
|
Report
|
Act
|
Reason
|
Waiting
|
What we
should do?
-Double click on
it to find the following details.
(a)Process start time
(b) Report/Program/Memory used
- Same like
SM50, but display global active W/p overview.
SM04:- Active User List
-It will
displays
|
Client
|
User
|
First name
|
Transaction
|
Time
|
Session
|
Type
|
Server
|
Type
may be
1.
Communication User.
2.
System user .
3.
Service user
4.
Dialog user
5.
Reference user
What
we should do?
-It
shows active user
-We
can kill the user. (Click user then system à end user)
-we
can kill the user session or complete user session
*AL08: - Global Active
Workprocess Overview
-It will
displays
|
Active instance
|
No of active users
|
No of interact user
|
No of RFC user
|
Then
|
Server name
|
Mand
|
User
|
Terminal
|
TCode
|
What
we should do?
-Check
the active user
-Can’t
kill user session
*SM21: - System logs
-It will
displays
|
Time
|
Ty
|
Nr
|
Cl
|
User
|
TCode
|
Mno
|
Text
|
Date
|
-Select
Date and time
-run
time problems are recorded
What
we should do?
Ø Check the text for getting the errors and problems
Ø Double click on it
Ø It will shows programatical errors
Ø user locking, Table space overflow
Ø Max extends reach
Ø update inactive, enqueue can’t issue lock
Ø W/p in priv mode
Ø most of dumps, buffer over flow
Ø all the d/b related errors, ex:- Ora 600, 1578,
1631,1632,1653,1654
Ø we can see local and remote systems logs also
Ø Here all system messages are recorded. Time, type of
message, client, user and all information will be available here.
Ø Here we can see local & remote system logs also
Ø Detailed list by double clicking on the message
SM37: -
Monitoring the Backgroud Jobs
Select job name, user name and execute
-It will displays
|
Job
|
Clnt
|
Job created by
|
Status
|
Start date
|
Start time
|
Duration
|
Delay
|
-
Status may be
1. Scheduled
2. Released
3. Ready
4. Active
5. Finished
6. Cancelled
What
we should do?
-Select job
and click job log
-it will
show date, time, message text, class no, type, Error occurs and reason
-our job is
to monitor jobs which are in cancelled status and unsuccessfully finished job.
Steps to follow the reasons for background job is not
running:
1.
Authorizations are
not assign to the user to run the particular job
2.
File which is to
be uploaded is not found
3.
We are running BDC
program from flat file to Database if the file structure is difference then job
is terminated.
4.
Table space
overflow
5.
Maximum extends
reached.
SM13: - Monitoring the terminated records
-It will
displays
|
Clnt
|
User
|
Date
|
Time
|
Shared Table
|
Lock argument
|
Status
|
Status may be
(1) init : the record is waiting to be updated
(2).auto :when the update task is activated from inactive
state then record will be updated automatically
(3).Run : processing
(4). Error: occurred that cause the request to be
cancelled
What we
should do?
check the record which are terminated
check types of updates
update modes
v1-critical update,v2-non critical update
synchronous :update d/b directly
asynchronous: update temporarily and then d/b
The following r the reasons to terminated
- There
are too many locks obtained by the users, the users are not getting the
locks to update his request
- Reaching
the maximum size of table, which can be identified by analyzing the sytem
lock and DB locks. One it is table size is fixed or auditing you need to
restart the update process manually
- If it
is frequently caused programming error this an be identified in SM13 and
inform the respective developer and the user.
SM14: - Update Administration
-it shows
Update/server/server group/parameter
and
Active/deactive
switch
What we
should do ?
v check
update is active or deactive
v if it is
inative explore the reasons for it
v we can
switch over between two, when error occurred in update process to make the d/b
consistent
v check
syslog and SM21 to get the information of update process error, and
inconsistency then deactive the update process by using parameter
RZ04: - Configuring Operation modes
|
Productive instances and their WP
distribution
|
|
|
|
Host Name Server Name Instance Profile
|
|
OP Mode Dia BP BPA Spo Upd Up2 Enq Sum
|
Maintain operation modes
(peak hours, off peak hours) for all the instances
What
we should do?
Ø Check whether the switching of operation modes is
performing or not
Ø if it is not performing check SM21
Ø
We can switch manually the w/p (dialogue and b/g
) depends upon peak and non peak hours
Ø go to SM63 and assign the time intervals for operation
modes
RZ20:- CCMS Monitoring
-COMPUTING
CENTRE MONITORING SYSTEM
-BUSINESS
PROCESS MONITORING (PERFORMED)
-
used to configure
the alert in the system
-
monitor set :
group of relative alert into for monitoring
-
monitor element :
this is an element under monitor sale which need to be monitoring
-
each element
consist of a certain threshold value
-
value reach the
threshold value alert will release
-
it will shows
alert text
-
O/s monitoring R/3
monitoring, D/b monitoring are performed here.
SPOOL MANAGEMENT
SAP provides print mechanism by using spool process.
Whenever the user request prints a document the dialog process handling the
request and stores the data TemSe(Temproary sequential objects). TemSe is
located either in database or in operating system(global directory), which will
determine by the profile parameter rspo/storage_location values will be g or gb
Access methods:
It
defines the type of printing method used by R/3 system. Access method is
specifies using protocol by three types:
- Local
printing
- Remote
printing
- Front-end
printing
Local
Printing: spool server transfers the data directly to the host printer or the
print manager. This is fastest printing is enable. C-windows L-Unix is the
protocols are used.
Remote
Printing: spool processor format the spool request data any spool request
information (no. of authors, no. of copies) and generates output request to a
different spooler. We use access method U-Unix; S-windows NT is the protocols
will be used.
Front-end
printing: user can print from request to printer directly
It is one-way
spool work processes are configured. It can be defined as either logical server
or real spool server 1. real spool server
2. Logical spools server
Real spool workprocess contains R/3 spool workprocess
Logical spool servers do not exist but interm it’s pointed
to real spool server. These are used to switch b/n real spool server for logon
load balancing and in place of failover.
There are various statuses of spool requests like waiting,
process, printing, problems, error, archievel
Various problems in spool request:
1. Devices not reached
2. Due to network failure
3. Printer problems
4. There is long queue at printer. If you are specified
sequential processing while the max. No. of spool request in the system 32,000
or 99,000
No comments:
Post a Comment