on 01-07-2016 10:11 PM
Hi All,
I have noticed that in all our NW 7.4 systems (different modules, different patch levels), whether they were installed as 7.4, or upgrade to 7.4, we see BTC workprocess stopping. It doesn't really seem to affect anything, as they start right backup and are not being used at the time....The trace files show the WP has reached the abap/heaplimit (set to recommendation)..... Has something changed in 7.4 with how the BTC Workprocesses are handled? Again, no affect, just see these messages over and over again.
Hi Diana,
Please supply the dev_wxx trace log with values for heap parameter.
abap/heap_area_nondia
abap/heap_area_total
abap/heaplimit
Regards,
Prithviraj.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi Prithviraj,
abap/heap_area_nondia = 2000000000
abap/heap_area_total = 7819755520
abap/heaplimit = 104857600
Thu Dec 31 10:30:53 2015
WP has reached abap/heaplimit = 52428800 bytes
Thu Dec 31 10:35:58 2015
ThWpNeedsRestart: abap strategy == kill, restart
ThWpRestart: restart wp (pid=59799) automatically
Disconnecting from ALL connections:
Wp Hdl ConName ConId ConState TX BC OC HC PRM RCT INF TIM MAX OPT Date Time DBHost Program
033 000 R/3 000000000 INACTIVE NO NO 000 NO YES NO NO 000 255 255 20151230 153740 dtc-sapbip
033 001 R/3*SID 033000004 DISCONNECTED NO NO 000 NO NO NO NO 004 255 255 20151231 101114 dtc-sapbip SAPLRRBA
033 002 R/3*BWMON 033000006 DISCONNECTED NO NO 000 NO NO NO NO 004 255 255 20151231 101114 dtc-sapbip CL_SQL_C
033 003 R/3*DTPLOG 033000005 DISCONNECTED NO NO 000 NO NO NO NO 000 255 255 20151231 102759 dtc-sapbip CL_RSBM_
Disconnecting from connection 0 ...
Close user session usr=706ce70 of con=0, svc=7038d48
Disconnected (con=0) from ORACLE.
Disconnected from connection 0
statistics db_con_commit (com_total=181, com_tx=130)
statistics db_con_rollback (roll_total=0, roll_tx=0)
***LOG Q02=> wp_halt, WPStop (Workp. 33 59799) [dpuxtool.c 318]
Thanks,
Diana
hi Diana,
The work process was restarted after using 50 MB out of ~2 GB of heap, this means there were other processes that were consuming heap memory.
If this on regular basis than please check if there is possibility to change the job timings to avoid memory bottleneck.
Else you can use RSMEMORY report to change the non-dia heap memory allocation without restarting the sap system to see if that helps.
Regards,
Prithviraj
Hello,
as describe in note 1571845 this is normal behaviour to release the heap memory that only gets released when the wp re-starts.
This is not an error.
You could increase "abap/heaplimit" parameter for instance to 100MB just not to so many messages in the system log.
This is kind of safeguarding parameter.
If you don't get memory dumps or memory issues you don't need to change the other parameter settings.
kind regards,
Mercedes
Dear Diana,
You dont have to do anything,
While executing anything work process allocated the memory until the execution is completed.
Once the activity is completed , it takes a restart to relaese the memory.
You need to analyze the program/job which is utilizing the memory and tune it accordingly.
With Regards
Ashutosh Chaturvedi
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hello Diana,
the explanation why the workprocess re-start with the message "WP has reached abap/heaplimit" could be found in SAP Note "1571845 - Error WP has reached abap/heaplimit- What does it mean?"
System resource usage (both memory and CPU) has increased with 7.4x kernel, so you could increase this parameter (double it) and see if the situation improves.
kind regards,
Mercedes
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
84 | |
10 | |
10 | |
10 | |
7 | |
6 | |
6 | |
5 | |
4 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.