BI Statistics Question

All,
It appears as though any query that we publish to portal that prompts the users to select variable values do not get logged in BI statistics. All queries without the variables are getting logged.
Is this the case for anyone else? Does anyone know why this could be the case?
Thanks
Rodney

Hi Nag,
All the objects initially activated have a default status D. You could change the default settings via, Extras --> Change Default. You can set the property of collecting statistics for the perticular object ON and OFF. Defauly for query is statistics collection OFF.
All the objects are treated as per the mentioned status except the queries, if the queries has default value "D" set, the system first determines the property for the InfoProvider of the query. If the InfoProvider has the value "D" system reverts to the default setting for the query otherwise it uses the InfoProvider setting.
For example, when a new query is created it has setting "D". The default setting for the query is Statistics Off. But if the InfoProvider for the query, however has the setting Statistics On. In this particular case, the system records statistics data for the query.
So in your case, if you know how the query will behave when the Default property is set, you first need to check what are the properties of its infoprovider.
Please let know if you need nay more information.
Regards,
Pankaj

Similar Messages

Optimizer Statistics questions

Hi All,
I have no clear picture about Optimizer Statistics
From Oracle 9i docs ( http://download.oracle.com/docs/cd/B10501_01/server.920/a96533/stats.htm#26713 )
GATHER_INDEX_STATS - Index statistics
GATHER_TABLE_STATS - Table, column, and index statistics
GATHER_SCHEMA_STATS - Statistics for all objects in a schema
GATHER_DATABASE_STATS - Statistics for all objects in a database
GATHER_SYSTEM_STATS - CPU and I/O statistics for the systemMy questions are:
1)if I use the 3rd option it should gather stats for all the objects including indexes. However, when I used that option it didn't gather satas for Index. Do I need to use cascade=>TRUE option?
2) I need to backup my current stats and gather new stats for entire DB. If I export DATABASE stats will it export all stats including index, system?
3) If I collect stats using GATHER_DATABASE_STATS, Do I need to explicitly collect other stats for index, system etc?
Any suggestions are most welcome.
Thanks,

Hi
1)if I use the 3rd option it should gather stats for
all the objects including indexes. However, when I
used that option it didn't gather satas for Index. Do
I need to use cascade=>TRUE option?
Yes, default is FALSE, so need to force TRUE
2) I need to backup my current stats and gather new
stats for entire DB. If I export DATABASE stats will
it export all stats including index, system?
Yes
3) If I collect stats using GATHER_DATABASE_STATS, Do
I need to explicitly collect other stats for index,
system etc?
No

Statistics questions

Hi, please excuse my ignorance, I have no experience with Discoverer, and my current client is using it and has thrown some questions/issues at me.
I also may add, I do not have access to my clients environment, not can I find an internal environment at Oracle to access.
Version: 10.1.2.54.25
They are reviewing data in table EUL5_QPP_STATS to identify problem workbooks/queries, etc.
However, they data they see is inconsistent, so they are questioning it's entire validity.
So, permit me if I may to seek answers.
1. Is there a document anywhere on the full description/purpose of the columns in EUL5_QPP_STATS (sorry the Discoverer Manuals on this subject are not helpful)
2. What could cause QS_ACT_ELAP_TIME to have a value of 0 (zero), yet QP_ACT_CPU_TIME have a value > 0 ?
3. What could cause the degree of precision for the values in QS_ACT_ELAP_TIME. eg: they noticed that all docs owned by one user has decimal values, yet all other rows not owned by this user have whole numbers for the value of this column ? (maybe related to question 2)
4. What could cause QP_ACT_CPU_TIME to have a null value ?
5. What does QS_CREATED_DATE represent ? The time the report/workbook began, or the time the record was created in the table, and when would that be? Upon report termination?
6. It is my understanding that you (Discoverer Admin) can set a role/responsibility so that it has Query Prediction and/or statistics enabled.
I cannot find documentation on precisely where this is set. So, in an effort to expedite the verification of these setting, is there a table.column we could run a query on to validate if these have been set accordingly ?
Thank You.

Hi,
I'll try and answer some of your questions:
1. Is there a document anywhere on the full description/purpose of the columns in EUL5_QPP_STATS (sorry the Discoverer Manuals on this subject are not helpful)No.
2. What could cause QS_ACT_ELAP_TIME to have a value of 0 (zero), yet QP_ACT_CPU_TIME have a value > 0 ? Maybe the elapse time is only calculated when the report finishes and if it fails to finish successfully there is no elapse time.
3. What could cause the degree of precision for the values in QS_ACT_ELAP_TIME. eg: they noticed that all docs owned by one user has decimal values, yet all other rows not owned by this user have whole numbers for the value of this column ? (maybe related to question 2)Sorry, no idea about this one.
4. What could cause QP_ACT_CPU_TIME to have a null value ?Maybe the report was created but never run.
5. What does QS_CREATED_DATE represent ? The time the report/workbook began, or the time the record was created in the table, and when would that be? Upon report termination? My understanding is that this is when the records are created, which is when the user successfully closes the connection.
6. It is my understanding that you (Discoverer Admin) can set a role/responsibility so that it has Query Prediction and/or statistics enabled. I cannot find documentation on precisely where this is set. So, in an effort to expedite the verification of these setting, is there a table.column we could run a query on to validate if these have been set accordingly ?The Discoverer preferences in the apps server prefs.txt (or in the registry if you are using desktop) control whether Discoverer collects stats and also whether these stats are used for query predition. On top of this you can also must set the stats collection privilege for a role/responsibility and this determines whether Discoverer store stats in the qpp stats table for that user/role/responsibility.
Rod West

InfoProvider statistics question?

Hello,
We are using BW release 3.5 level 0020.
What is the difference between 'Tools => BW Statistics for InfoProviders' and
'InfoProvider => Manage => Performance tab => Create Statistics?
I have selected 'Refresh DB statistics after each data load'. A job beginning with BI_STAT runs after each extraction into the InfoProvider.
I would like to schedule this job under DB Statistics => Create Statistics, but I noticed that the Job name begins with BI_INDX.
Is this the same as the BI_STAT job?
Thanks,
Mark

Mark,
[i]BW Statistics for InfoProviders'
Thsi is technical content - implying system collecte stats abt infoproviders..such as last accessed etc..USed for performance tuning etc
'InfoProvider => Manage => Performance tab => Create Statistics?
This is DB statistics: this would be utilized by system to create query run path to optimize query access path... etc..
Hope it helps
Sorabh

BW Statistics: my record time vs time in BW Statistics

Hi all,
Anyone knows or have documentation of how it works BW Statistics? I mean, from the moment it start recording the time in initializing the OLAP processor, the time when the user can see the page, to the time to end the load of the page or query. I already have seen the documentation in
http://help.sap.com/saphelp_bw30b/helpdata/en/f2/e81c3b85e6e939e10000000a11402f/frameset.htm
and the notes:
- 309955 - BW statistics - Questions, answers, errors
- 750554 - Importance of time stamps in RSDDSTAT
- 608483 - BW statistics: Reports for Web reports
My problem.
I have some test, I have been recording the times to take the load of a query and after I compare with the time that shows the model BW Statistics and it not the same! I've been thinking that these can be possible because the load of a query not ends when the page is show to the user, that is because the query is still doing pagination, preparing buffer, loading images, all that I suppose it, so the question is where is the registry that have the time when the user can see the page?? where is the record of that time?
I'm in BW 3.5.
I record the load of a page with the program HttpWatch.
I hope you can explain to me where I can get that registry of time.
Any input is greatly appreciated.

I was looking for the same KFs, but did not find anything meaningful in ST03N as well.
But I would recommend to execute the Web Report (from WAD or from Portal, if you are using Enterprise Portal) and then look at table view RSDDSTAT_OLAP. In this table view, BW Statistic data for each step is saved on event level. So if you execute the Web Report, you will find entries for the step corresponding to displaying the selection screen and for the step of generating the Web Report after the selection screen. The column RUNTIME displays the runtime of each step, which is a good indicator for the average runtime KFs you want to have.
Maybe there are more meaningful KFs in technical BI Content, but I did not find them yet.

BI Statistics cannot be captured in Portal or Query Designer

We're experiencing inconsistent behavior of the BI Statistics capturing mechanism, possibly after our recent upgrade to SAPKW70015. It appears that only queries run via the Bex Analyzer are correctly captured in log tables (RSDDSTAT), whereas queries run via the Portal are not (instead, we get unknown entries with object name RSDRI, rather than query/infoprovider names in the log tables).
We've developed a very sophisticated application to report on BI usage ( top n queries by infoarea, by infoprovider, by CPU, by DB-records, by OLAP-records, all queries used by a user etc) and to find out who is using which query, as well as associated resource consumption, at any given time, in real time.
This bug completely throws us off balance. We can no longer monitor query acitivites on the Portal; nor can we trust the BI statistics we've collected in the last two years. Another BI installation has reported the same problem to SAP. SAP seems to think upgrade to SP17, which could be a major step, will fix the problem. Is there anything we can do in the mean time to fix this problem without having to wait for upgrade to SP17 ? Is there an OSS note on this issue ?
(we have SAP_BW Release 700 SP15; SAP_ABA/SAP_BASIS Release 700 Level 13; SEM-BW Releae 602 Level 1; ST-PI 2005_1_700 level 5; PI_BASIS 2006_1_700 level 2; ST-A/PI 01H_BC0700 level 0)
Suggestions will be very much appreciated.
Henry Hone.

Hi,
If the BW statistics are not recorded for a Web Report, this may be because recording was not activated for at least one query (or its InfoProvider). Depending on the calculation sequence, the BW statistics may also not be written for the other queries. The solution is to activate the BW statistics for all of the Web Report InfoProviders.
Check oss notes
Note 608483 - BW statistics: Reports for Web reports
Note 309955 - BW statistics - Questions, answers and errors
Hope it helps.
Thanks
Mona

What is the best approach to install BI statistics in SAP BI ?

Hello All,
what is the best approach to install BI statistics in SAP BI ?
by collecting objects in standard BI content- 0TCT*objects or
by executing some standard tcodes.
Regards,
Siva

the best approach is based on version of your BW system follow up install steps in notes:
BW 3.x:
309955 - BW statistics - Questions, answers and errors
BW 7.x
934848 - Collective note: (FAQ) BI Administration CockpitBW 7.x
Cheers,
m./

Bug in Bw statistics

Hi gurus,
I have a strange problem in BW Statistics.
As you know the key of RSDDSTAT table is STATUID, and for each navigation should be one entry in this table.
The problem is that in BW Statistics query I got the number of navigations, for one STATUID value, is more than 1. In addition, the other statistics values in BW Statistic cube are not equal to the values in RSDDSTAT table. e.g QOLAPTIME, Selected DB etc.
I think this problem occurs just when I run a query in WAD.
Is there a note which describe this problem ?
Thankfully,

hi,
hope following oss note may explain this ....
- 309955 - BW statistics - Questions, answers, errors
- 750554 - Importance of time stamps in RSDDSTAT
- 608483 - BW statistics: Reports for Web reports
309955
a) The database access times (table RSDDSTAT) are not always recorded correctly, but added to the OLAP time.
750554
... In general however you will not always be able to easily compare the difference in time stamps with the specified times of a data record.

Alter index rebuild question

Hello experts,
I have a simple question regarding rebuilding indexes...
First, I am running 11.2.0.2 on Solaris 10.
My reason for rebuilding is to relocate the indexes to their own tablespace, and while I recognize this is a subjective reason as far as performance is concerned, it makes administration easier.
That being said, my question is in regards to use of "compute statistics" and also "online", and "nologging".
Since I do want to minimize he impact to our users, and this table is one of the most used tables in the database, my syntax was going to be as follows:
alter index <schema>.<index> rebuild online nologging tablespace <ts_idx> compute statistics;
Questions:
1. does doing the rebuild online preclude queries against the table during the duration of the rebuild?
2. does use of nologging improve speed of rebuilding
3. is it advisable to include the "compute statistics". I believe I read on Jonathan Lewis' page that statistics are gathered automatically and to include "compute statistics" was not necessary when rebuilding an index.
And I guess one final question is whether I can include all the options in the same rebuild statement as I have shown above?
I've tested the above in our test database, but it seems to hang when using the online option (with all the other options).
Thanks in advance.

Excellent resource. Thanks for the link. While I've read most of those discussions on either Jonathan's or Richard's web sites already, it is a good presentation and reference.
Unfortunately, it does not answer either of the questions that I asked. The reason I asked those questions is because the answers you gave contradict what i read in Oracle documentation (although it might be version specific).
Also, it does seem to be slightly faster to rebuild using nologging (in my test environment).
I've tested this with multiple indexes and the timing is slightly improved consistently.
{quote}
SQL> ALTER INDEX ARADMIN."T2179_C1_0944_1037" REBUILD LOGGING TABLESPACE REMEDY_IDX;
Index altered.
Elapsed: 00:00:10.68
SQL> ALTER INDEX ARADMIN."T2179_C1_0944_1037" REBUILD nologging TABLESPACE REMEDY_IDX;
Index altered.
Elapsed: 00:00:08.50
{quote}
The use of online, however, seems to "hang" for some reason, even in my test database while no one else is logged into the database except me.
Referencing Oracle's documentation, it says that rebuilding an index online does not allow queries against the base table. This is why I asked this question, but your answer is different than Oracle's. In this case, however, it is for Oracle 10.2, so not sure if this is still true for 11.2.
Ref. Oracle Text SQL Statements and Operators
[ONLINE]
ONLINE enables you to continue to perform updates, inserts, and deletes on a base table; it does not enable you to query the base table.
You cannot use PARALLEL with ONLINE. ONLINE is only supported for CONTEXT indexes.
Now, regarding the use of "compute statistics", I am pretty sure I recall reading either in Jonathan's papers, or in Richard's papers, that this is not necessary and/or is redundant. But then why would Oracle include this as an option. This is counter intuitive, which again is why I'm asking. Do you know if any reference that can clear this up?

0BWTC_C08 cube, virtual provider Query Problem

Hi Gurus,
Has anyone worked on 0BWTC_C08 cube. it is statistics virtual provider. i have installed it in BW and ran a query on it without any selections, 0TCTIFCUBE in rows and frequency in columns it is returning null data for all the cubes.

Hi amarri9:
Take a look a t the SAP Note 309955 - "BW statistics - Questions, answers and errors".
Regards,
Francisco Milán.

Bw statistic - number of navigation

Hi gurus,
When I execute a new query, this should increase one of number of navigation. That's right... ?
In most cases it is, but sometimes I get number of navigations is 3 even 4 per one time executing.
Is there any suggestion why do it occurs ?
Thankfully,

hi,
not sure if it's set incorrectly
309955 BW statistics - Questions, answers, errors
Prior to Support Package 23 (2.0B), 15 (2. 1C)
a) Counter 0TCTNAVCTR is set incorrectly, if a Web report contains several queries where navigation is performed several times. The problem is solved with the specified support packages or Correction 386678.
Prior to Support Package 19 (2.0B), 11(2. 1C).
a) When using Web reports, key figure 0TCTNAVCTR is not set correctly.
                        The cause is that, when using Web reports, a large number of entries that cannot be assigned to any query (as explained above under 'Explanations') are generated in Table RSDDSTAT. The number in key figure 0TCTNAVCTR is only calculated as follows when loading the data. All records selected are sorted by the SESSIONUID (records of a session) and NAVSTEPUID (all records of a navigation step). When loading, the value of 0TCTNAVCTR is set to 1 each time the NAVSTEPUID is changed (otherwise 0). This way the number of navigation steps can be easily determined.
                        In the case of Web reports, however, records of handle 9999 (see above) are inserted in between so that the counter is set to 1 too often.
                        The specified correction 306437 solves the problem.

Question on gathering statistics in Enterprise Manager

When in EM, choose Server->Query Optimizer->Manage Optimizer Statistics->Gather Optimizer Statistics:
My question, when choosing the Object Type "Database": will this gather and optimize all of the other objects below database (i.e. Schemas,Tables,Indexes,etc.)? If I run the optimizer against a given schema, will this also compute statistics for all of the tables and indexes in the users schema?
I tried looking through the documentation set but it is not clear to me. Can someone break this down sesame street style for me?

934865 wrote:
When in EM, choose Server->Query Optimizer->Manage Optimizer Statistics->Gather Optimizer Statistics:
My question, when choosing the Object Type "Database": will this gather and optimize all of the other objects below database (i.e. Schemas,Tables,Indexes,etc.)? If I run the optimizer against a given schema, will this also compute statistics for all of the tables and indexes in the users schema?
yes
I tried looking through the documentation set but it is not clear to me. Can someone break this down sesame street style for me?http://www.oracle.com/pls/db111/search?word=dbms_stats&partno=

'Gather Schema Statistics' concurrent request question

Hello,
I just have a question about the 'Gather Schema Statistics' concurrent request.
We run this every evening on ALL schemas, however our SYS and SYSTEM schemas aren't registered in the application so therefore stats aren't gathered for these.
Should SYS and SYSTEM be registered and stats gathered for these 2 schemas every night?
Please advise
Thank you
Sarah

Hi Sarah,
First of all, Its not necessary to run gather schema statistics every evening. Either once in a week or month would suffice the need. Please look at the following note:
How Often Should Gather Schema Statistics Program be Run? [ID 168136.1]
With respect to your question:
Sys and system users are not included in the "ALL" specified parameters in gather schema statistic program. It includes the list of product related users. But it is recommended you collect statistics for sys and system users and you can automate and achieve this by creating a job and schedule it at DB level. Please follow the following note which will assist you:
EBPERF FAQ - Collecting Statistics in Oracle EBS 11i and R12 [ID 368252.1]
Cheers,
Asif

Error in Export "EXP-00091: Exporting questionable statistics "

when i execute export on server having RAC DB R10g R2 on Linux 4 it return error , nobody has changed it's characterset .
how to fix this error, CAN THIS ERROR CAUSE BIG DESASTER TO DB IN FUTURE
C:\>exp hst/hst file='/home/oracle/expdat.dmp
About to export specified users ...
. exporting pre-schema procedural objects and actions
. exporting foreign function library names for user NRI
. exporting PUBLIC type synonyms
. exporting private type synonyms
. exporting object type definitions for user NRI
About to export NRI's objects ...
. exporting database links
. exporting sequence numbers
. exporting cluster definitions
. about to export NRI's tables via Conventional Path ...
. . exporting table TABLEA 0 rows exported
EXP-00091: Exporting questionable statistics.
. . exporting table TABLEB 0 rows exported
EXP-00091: Exporting questionable statistics.
Kindly Help be ho can i fix this error
Thanks in advance

Hello,
You wrote:
CAN THIS ERROR CAUSE BIG DESASTER TO DB IN FUTUREThe error message EXP-00091: Exporting questionable statistics just warns you that the statistics
were exported but these statistics may be not useable.
So, if you have to import these Tables, you may have to collect again the statistics of these Tables.
It's not relative to this problem but why don't you use DataPump (expdp/impdp) ?
Please find enclosed a link about DataPump, it's much more powerful than the classical export/import:
[http://download.oracle.com/docs/cd/B19306_01/server.102/b14215/dp_overview.htm#i1009203]
Best regards,
Jean-Valentin
Edited by: Lubiez Jean-Valentin on Feb 6, 2010 12:56 PM

EXP-00091: Exporting questionable statistics

Dear Friend,
I am using Linux Database and Windows database, the problem now I am facing while exporting data from Linux which is imported from windows oracle database dump file.
when I am exporting I am getting this error
EXP-00091: Exporting questionable statistics
and shows
Export terminated successfully with warnings
Now exported file I am importing from Linux to windows and getting this error
Connected to: Oracle9i Enterprise Edition Release 9.2.0.1.0 - Production
With the Partitioning, OLAP and Oracle Data Mining options
JServer Release 9.2.0.1.0 - Production
Import file: EXPDAT.DMP > c:/aa.dmp
Enter insert buffer size (minimum is 8192) 30720>
IMP-00010: not a valid export file, header failed verification
IMP-00000: Import terminated unsuccessfully
Regards,
Oracle User

Hi,
>>because I exported dump from oracle10G
Then you use the exp 10g version, right ? To move data DOWN a version(s), you need to export using that lower versions EXP tool and using IMP that lower version tool.
Take a look at this below?
Exporting Data From Release 10.2 and Importing Into Earlier Releases
Export From      Import To      Use Export Utility For      Use Import Utility For
Release 10.2      Release 10.2      Release 10.2           Release 10.2
Release 10.2      Release 10.1      Release 10.1           Release 10.1
Release 10.2      Release 9.2      Release 9.2           Release 9.2
Release 10.2      Release 9.0.1      Release 9.0.1           Release 9.0.1
Release 10.2      Release 8.1.7      Release 8.1.7           Release 8.1.7
Release 10.2      Release 8.0.6      Release 8.0.6           Release 8.0.6
>>EXP-00091 is still remain.. you said I have to chang ethe NLS but I am new to linux
I think that there is no problem ... but you can use also as I said before the statistics=none clause, but remember that you need to use the exp 9i version ...
Cheers

BI Statistics Question

Similar Messages

Maybe you are looking for