Prod. system periodically into swap

Hi all
I have problem with dialog production instance. Periodically dialog going into swap and when swap is full accordingly instance is not responding.
Configuration:
ECC 5.0 Oracle 9.2 Solaris 10
I want to understand what options are actually used physical memory?
CI 16G RAM + 5.8G Oracle SGA:
ztta/roll_area 3000000 Byte Roll area per workprocess (total)
ztta/roll_first 1 Byte First amount of roll area used in a dialog WP
ztta/short_area 3200000 Byte Short area per workprocess
rdisp/ROLL_SHM 16384 8 kB Part of roll file in shared memory
rdisp/PG_SHM 8192 8 kB Part of paging file in shared memory
rdisp/PG_LOCAL 150 8 kB Paging buffer per workprocess
em/initial_size_MB 5000 MB Initial size of extended memory
em/blocksize_KB 4096 kB Size of one extended memory block
em/address_space_MB 4092 MB Address space reserved for ext. mem. (NT only)
ztta/roll_extension 2000000000 Byte Max. extended mem. per session (external mode)
abap/heap_area_dia 90000000 Byte Max. heap memory for dialog workprocesses
abap/heap_area_nondia 2000000000 Byte Max. heap memory for non-dialog workprocesses
abap/heap_area_total 2000000000 Byte Max. usable heap memory
abap/heaplimit 40000000 Byte Workprocess restart limit of heap memory
abap/use_paging 0 Paging for flat tables used (1) or not (0)
dialog 16G RAM:
ztta/roll_area 3000000 Byte Roll area per workprocess (total)
ztta/roll_first 1 Byte First amount of roll area used in a dialog WP
ztta/short_area 3200000 Byte Short area per workprocess
rdisp/ROLL_SHM 16384 8 kB Part of roll file in shared memory
rdisp/PG_SHM 8192 8 kB Part of paging file in shared memory
rdisp/PG_LOCAL 150 8 kB Paging buffer per workprocess
em/initial_size_MB 12000 MB Initial size of extended memory
em/blocksize_KB 4096 kB Size of one extended memory block
em/address_space_MB 4092 MB Address space reserved for ext. mem. (NT only)
ztta/roll_extension 2000000000 Byte Max. extended mem. per session (external mode)
abap/heap_area_dia 2000000000 Byte Max. heap memory for dialog workprocesses
abap/heap_area_nondia 2000000000 Byte Max. heap memory for non-dialog workprocesses
abap/heap_area_total 2000000000 Byte Max. usable heap memory
abap/heaplimit 40000000 Byte Workprocess restart limit of heap memory
abap/use_paging 0 Paging for flat tables used (1) or not (0)
In st02 on both prod. instance i have red swaps:
Field description, Initial records, CUA, Screen, Generic Key, Single record, Export/Import.
CI:
Hitratio % = 97.57
90.00
dialog:
Hitratio % = 98.73
87.00
The average number of users per day : CI: 47
dialog 58
Tell me please what could be the problem and how to solve it?

Andrey i can see your system has been configured incorrectly
Your total extended memory for the instance (em/initial_size_MB) is 5GB and maximum extended memory each work process can use is is 2 GB (ztta/roll_extension) . so incase you have 3 work process who want to use 2GB of extended memory you will be already exhausted because your total extended memory is 5 GB.
Below is my recommendation assuming you do not have any other instances on each of the CI and Dialog instance box.
em/initial_size_MB = 12 GB
ztta/roll_extension = 300~600 MB
Heap Area for dialog = 1.5 GB
Heap Area for Non diag = 4 GB
No of diag WP = 20
No of Btc WP = 15
Note : Our normal thought process is we have to multiply no of (diag+BTC) wp and ztta/roll_extension to arrive at a figure for
em/initial_size_MB but that is not the right way of calculating because not all the work process are going to use extended memory only few use them in case some memory consuming reports are ran.
Try them out and let me know. Keep a good luck on ST02 and ST03.
Cheers !
Manish Jain

Similar Messages

SWF file is not working in Portal PROD System

Hi All,
We have written custom mxml script in Adobe Flex 3 which displays the list of applications stored in the DB. Then we have posted that .swf file into our Portal KM and then we have created one URL iView with the location from KM.
These days we are facing issue like the output of the script is not working means the applications are not listed on the portal prod system.
Whenever the issue arieses we are requested to upload the file once again into KM and it works after the upload been done.
We are not sure what could be the reason for this issue and I request any one to assist us to resolve this issue.
Please let me know if you need more info on this issue.
Thanks in advance.
Best Regards,
Vijaykumar K.

Hi Vijay,
I belive this should be a portal issue, anyways let me give it a try.
I never worked on flex in SAP but did similar stuff in portal content development.
Do you have any traces in that development..?
Did you check any messages being loged in portal...?
Check if there is any issue with the file like couldn't recognise a resource from KM which is required etc.
Cheers,
Sai

Delete orphaned objects in Prod system

Hi all,
When our BW landscape was set up years ago, some test source systems were created (along with DataSources etc) in the BW DEV system and migrated into QAS and PROD. The source system in DEV was deleted long ago, but there are remnants still left in our PROD system.
Since PROD is locked to any changes and Source Systems are not transportable anyways, I believe the only way to delete the objects in PROD is to have our BASIS team unlock the system and then delete the source system which should also delete any objects associated with it.
Just wondering if there any other options to remove the old objects from PROD, or if opening the system and deleting is the best way to go? Many of those were created as 3.x versions, and I am preparing for an eventual upgrade to our 7.1 system. Therefore, I am trying to eliminate or update any of these old 3.x objects to prevent any issues during upgrade.
Thanks!
Darryl

Hi Erica,
The best option to make the production and quality system in sync is to copy the Production system into the Quality system. Your basis team can do this. In this case, both the Production and the Quality system will be in sync and the development system will remain the same to update the the quality system at some later stage.
Hope this helps...

How to compare all macros in the Macro workbench with DeV & Prod system

I have one issue with macros, all the macros in maco workbebnch in developement system are not in sink with Production system.
I want to campare macros in both the system & based on comparison need to add or deactivate macros in the Developement system & move Transport request to accross the landscape like Dev-QA& Prod system.so that all macros are in sink with three different system.
Can anybody advice,Is there any tool or T-code that compare all the macros in the two different systems.

This code is different from yours. For one thing, the first part, where inOrder is false, is not included because you did not post what behavior that is supposed to perform. The code below does what you posted...it puts some Strings into a Stack, then parses the Stack. If adj is followed by NP* then adj is removed. Then the new Stack is parsed to find det followed by NP*, and det is removed while NP* is changed to NP.
import java.util.Stack;
import java.util.Vector;
public class Test {
 public static void main(String[] args) {
 Stack stack=new Stack();
 Stack stack2=new Stack();
 String[] s={"no","adj","NP*","adj","not","NP*","adj","NP*","dumb","det","NP*"};
 for(int i=s.length-1;i>-1;i--) stack.push(s);
stack=structureCheck(stack,"adj","NP*","NP*");
for(int i=stack.size()-1;i>-1;i--)
System.out.println(stack.elementAt(i));
System.out.println();
stack=structureCheck(stack,"det","NP*","NP");
for(int i=stack.size()-1;i>-1;i--)
System.out.println(stack.elementAt(i));
public static Stack structureCheck(Stack s,String tofind,String tomatch,String tochangeto) {
Stack orderStack = new Stack();
for(int i=0;i<s.size();i++) {
String str=(String) s.elementAt(i);
if(i<s.size()-1 && str.equals(tomatch) && ((String) s.elementAt(i+1)).equals(tofind)) {
i++;
orderStack.add(tochangeto);
} else
orderStack.add(str);
return orderStack;

Error while generating export datasource in Test and Prod system

Hi experts,
I tried to generate export datasorce on a cube in teh testing system so that i can use this as a datasource for another cube. I did this so that i can transport all teh relevnt objects from D to T. But Iam gettin teh following erro msg
System setting does not allow changes to be made to object ISTD 80CRM_PRI.
while Prod system produces teh follwoing error msg
Changes to Repository or cross-client Customizing are not permitted
Choose 'Display object' or 'Cancel'.
After i have done this and when i go to the context menu of teh infocube i can see maintain export datasource option.
Also I couldn't see the infosource generated in Infosources ( FYI I have already done display generated objects under settings)
Can anyone tell me how to proceed further.
Thanks,
DV

Hi DV,
Generally export datasource are not included in the request from the test server. It gets generated automatically when (ODS,Cube) based on which the datasource s transported to production / quality .
I think solution should be try not to include export datasource, infosource ,transfer rules to request from dev server.
Hope that helps.
Regards
Mr Kapadia

Sol10 u8 installed on a ZFS Root File System have different swap needs?

Does Sol10 u8 installed on a ZFS Root File System have different swap needs/processes?
Information:
I've installed Solaris 10 (10/09 s10s_u8wos_08a SPARC, Assembled 16 September 2009) on a half dozen servers and every one of them no longer mount swap at boot.
The install program commented out the old swap entry and created this one:
# grep swap /etc/vfstab
swap - /tmp tmpfs - yes -
Everything works like a champ. I didn't discover the issue until I tried to install some patches and the install failed. It didn't fail because of lack of swap - it refused to run because it found "No swap devices configured".
Here are the symptoms:
# swap -s
total: 183216k bytes allocated + 23832k reserved = 207048k used, 13600032k available
# swap -l
No swap devices configured
# mount | grep swap
/etc/svc/volatile on swap read/write/setuid/devices/xattr/dev=5ac0001 on Mon Apr 19 08:06:45 2010
/tmp on swap read/write/setuid/devices/xattr/dev=5ac0002 on Mon Apr 19 08:07:40 2010
/var/run on swap read/write/setuid/devices/xattr/dev=5ac0003 on Mon Apr 19 08:07:40 2010
#

Hi Nitabills,
I assume that you create a zfs entry for swap with the commande zfs create -V $size
did you launch the command :
swap -a /dev/zvol/dsdk/$ZPOOL/swap
Try this entry below in the vfstab :
/dev/zvol/dsdk/$ZPOOL/swap - - swap - no -

How to remove Variable-Values prior to transporting to prod-system?

Hi specialists,
I've got the following problem:
We've got a mixed BEx-client-environment with user using BEx 3.5 as well as BEx 7.x.
So I create workbooks with BEx-Analyzer 3.5 and transport them from the developement-system to the production-system.
The problem: Obviously the variable-selection that I used at last in the developement-system is saved in the workbook. So let's say I tested the query in the dev-system with "DE01" as value for the company_code on the variable screen, then this value is saved in the workbook and transported to the prod-system.
In the production-system the value is not replaced anymore by the users permitted company-code values (the variable is based on the users permissions), but the user has to remove this "DE01" manually and has to enter all his personal company-codes manually - which makes the permission-bases-variable senseless.
Even if I clear all variable-values in the developement system (which is not possible for mandatory variables) prior to saving the workbook, the issue is the same: This time empty fields are presented in the prod-system, but the permission-based-variable isn't set correctly - the field stays empty (=exactly as it was when I saved the workbook in the dev-system).
So, how can I clear the variable-screen-values before I save and transport the workbook to the prod-system? The only way (and it is certainly not the recommended one) i found: If I cancel the query at once after having confirmed the variable-screen and save/transport that query, then the variables seem to be really empty and the permission-based variables in the prod-system work.

So, when user run the report in production, they will have to remove # for mandatory variables and put thier value.
Mh - most of our reports have variables for "company_code","sales_org" (for example).
Normaly this fields are not mandatory but optional.
But: Normaly if the user opens such a variable-screen (for the first time) the company_code and sales_org are preset according to the users permissions for this chars.
If I follow your description, then the variable-screen would open with "#" or "" - but wouldn't be replaced by the user's permitted values anymore :/ - hence the phone rings because the users didn't manage to restrict the variables to their permitted values.
Maybe this is a case for an oss-message for the area BW-BEX-ET-USAB Userability Frontend....

How to convert system Date into DD-MMM-YYYY Format

hi friends,
please help me How to convert system Date into DD-MMM-YYYY Format.
Regards
Yogesh

HI..
data: w_dt(11) type c,
         w_month(2),
         w_mon(3).
w_month = p_date+4(2). **p_date = given date**
case w_month.
when '01'.
w_mon = 'Jan'.
when '02'.
w_mon = 'Feb'.
when '03'.
w_mon = 'Mar'.
when '04'.
w_mon = 'Apr'.
when '05'.
w_mon = 'May'.
when '06'.
w_mon = 'Jun'.
when '07'.
w_mon = 'Jul'.
when '08'.
w_mon = 'Aug'.
when '09'.
w_mon = 'Sep'.
when '10'.
w_mon = 'Oct'.
when '11'.
w_mon = 'Nov'.
when '12'.
w_mon = 'Dec'.
endcase.
Now...
concatenate p_date6(2) '-' w_mon '-'   p_date0(4) into w_dt.
write w_dt.

Consistency between dev and prod systems!!!!

Hi All,
I wuld like to check the consistency between Dev and Prod systems for all the Data model and other objects existing. I would like to do this sanity check to see whether the two systems are in sync. Are there any tools or transaction codes within SAP framework which can guide me in this direction.
Kind Regards,
Surya Tamada.

Hi Surya ,
First build whole data flow :
1 .Create Infoarea ,Infoobjects & Masterdata ,Infoproviders ,Infoobject Catelog under Infoarea,MetaData - Characteristics MeteData - Key Figures
2 DSOs ,Infocubes ,Infosets, Multiproviders ,Views in the development system ,Datasources for BW system,Flat File Datasources,Transformations
then get data from the se11 tables as per your requirement :
like RSDVCHA ,RSDBCHATR,RSDCUBEIOBJ,RSDICMULTIIOBJ etc and built view on it as per your requirement .
then built flat file data sources and transformation .
get all data downloaded to flat file in ur PC .In the same way from other system.
make flow from PC File Datasource -> DSO -> Infocube and build queries for different objects .load the downloaded files to them and then run query .
ex : query to compare cube will take cube name and systems name as input and compare objects .
Regards,
Jaya

Source System Mapping between BW Dev and Prod system

Hi,
I have a DB Connect Source System in BW Dev system called DBSRCDEV which points to one of our Oracle Dev DB.
In Production, the DB Connect Source System is called DBSRCPROD and it points to the corresponding Oracle Prod DB.
In Development, I created a data source on this Source System (DBSRCDEV) and I tried to transport it to Production. But it failed with error saying that DBSRCDEV does not exist.
Where and how can I maintain or configure the mapping between DBSRCDEV and DBSRCPROD Source System so that I can transport my data source to Production?
Please help.
Thanks,
CH

Hi,
usually you need to maintain the setting for the change of the logical system name (source system) in each target system of a transport. So in your prod system goto rsa1->tools->change logical system names...
After maintaining that table, try to import your request again.
regards
Siggi

Integration Directory (ID) transport to Prod system

I am trying to move XI ID and IR to prod system thru file system transport. When I transport ID requests and activate it keeps complaining "Unable to move object. Change lists were not part of the same import"
I checked the ID there are no open requests in my id or others all have been closed.
I would also like to know if there is a way to completely transport the Integration Directory (ID) objects either scenario or service objects along with their change lists
Thanks
Subbu

hi,
Change lists were not part of the same import"
Make sure you have activated change lists open under ur id and check if any for other userid
 would also like to know if there is a way to completely transport the Integration Directory (ID) objects either scenario or service objects along with their change lists
Once you have activated the changelists you can either transport the whole scenario(right click on scenario and export) or transport individual objects using the tools-> export option on the tool bar of ID.
Cheers,
naveen

Manual billing for legacy system period Billings

Hi Gurus,
Need your help.
My Client wants to do Billing for Legacy system period periods. Is it possible? Can we do through Manual Bill? If yes how SAP system can accept those contracts or old dates from Legacy?? I know that we can do Mnaul billing through T-code:EA16.
appreciate earliest response.
thanks
DGR

Hi,
Yes we can do it.There is one field in Ever called XVERA which signifies legacy move-in.If u migrate data from Legacy system and maintain that field,then only u can create manual bill .i have done the same as my users required that.just try this, i can help u for the same even.
Regards,
Raman

Convert system time into User local time.

Hi,
How to convert System time into User Local time.
Are there any FMs available.
Regards,
Kiran

hi,
There is just one simple funda behind the Date and time logic in SAP.
-> It is absolutely useless to store Date and time directly in DB. Why ?
Because 00:00 a.m in London could mean 05:30 in India !! So If I (in India) store a Date 1st Feb 2008 , 10:00 a.m .. This time for a person in London could be invalid.. because he is behind your time zone..
So the solution is to store the date/times in Timestamps format. Taking the above example.. 1st Feb 2008, 10:00 a.m (in India) will be stored as 20080201043000.
This time if seen in German Time zone will look as : 1st Feb 2008, 05:30 a.m.
Simple right
-> In India we store date as DD.MM.YYYY while in other countries they store it as MM.DD.YYYY.
The separator could be . or - or / .
The date format is user dependent. I might prefer to store it as DD.MM.YYYY and you might want to store it as MM-DD-YYYY.
System does not bother what your format is. It always stores in YYYYMMDD (8 Char)
On screens if you want to display Date then it must always be shown in that users format who is seeing it. How to get this done ?
DATA : lv_date(10) type c.
write sy-datum into lv_date.
So in LV_DATE variable the date will be stores in the users format.
Simple right
There are function modules that do the same. But normally we should avoid use of FM to make code simpler, short and easy... no complications at all ..
You can use system variables sy-datlo (Local Date for Current User) and sy-timlo ( Local Time of Current User ).
You can also use GET TIME statement.
reward if useful,
preet

When the system goes into sleep mode it keeps making a beeping noise. The system is shut down by holding the button in

When the system goes into sleep mode it keeps making a beeping noise. The system is shut down by holding the button in
and will then boot up again without any errors.
I Have replace RAMS, Motherboard and Power Supply. I have check the Bio's and Power setting on the unit. there is no place to cause this issue.
Please help
PC Power Tech

Syamak,
You didn’t specify where the beeping noise is coming from.
If it’s coming through your speakers, you could try to update your audio drivers since it’s possible they’re incorrectly sending a signal to the speakers as the audio chipset enters standby.
If the noise is coming from the motherboard, you’ll need to take a closer look at your system’s configuration or contact the manufacturer.
I hope this helps.
Mike
Windows Outreach Team – IT Pro

How DAC decodes its Source System Parameters into Informatica mappings??

Hi All,
I would like to know about "How DAC decodes its Source System Parameters into Informatica mappings??".
Is it an informatica mapping? or through a file? or its internal to DAC server?
Can we say the DAC server to decode it differently i.e. can we costomize the decode process? For example
$$START_DATE - i gave "Custom Format(1950-05-21 01.23.56) @MM/DD/YYYY" at dac - and it decodes as '05/21/1950' but I want it to decode as TO_DATE('05/21/1950','MM/DD/YYYY') is this possible?
Note that the informatica mapping takes the first decoding '05/21/1950' as number and through a sql error.....
Thanks for your time......
DK

Please set the values of $$FILTER_BY_SET_OF_BOOKS_ID, $$SET_OF_BOOKS_ID_LIST in DAC and check if the values get reflected in the parameter files that DAC generates for informatica at run time.
If this does'nt work, see if the BI services were restarted after doing change in the parameter file. The services need to be restarted in order to reflect the change.

Prod. system periodically into swap

Similar Messages

Maybe you are looking for