STATSPACK/AWR

Similar Questions

• Troubleshoot the interconnection

  Hello world... If there is a node 3 CARS and interconnection does not work... What are the steps to solve the interconnection... - What is the job of the DBA or network admin guy job?.

  If interconnection does not at all because ping fails it's a network problem that must be fixed by the network administrator.
  If you think that interconnection is too slow, it's something that can be checked with:
  1 ifconfig - a or netstat s
  2. with the timed 5 Albums of Statspack/AWR
  Please read page 20 to 23 in CAR Performance Experts reveal all the http://ioug.itconvergence.com/pls/apex/CILOUG.download_my_file?p_file=382

  Edited by: P. Forstmann on 28 Dec. 2010 17:09

• Cancel in memory

  I'm in the document provided by Craig-
  http://www.Oracle.com/education/america_eblasts/FY09/Q1/craig_shallahamer_all_about_oracle_s_in_memory_undo.PDF
  
  and I have the following queries-
  
  1. I can tell of this document that the IMU reduced significantly different types of overhead costs. What I can't say is what the cost is to achieve? If it is in memory, it uses memory. How much memory is consumed for a database by using this feature? Are there other 'costs' associated with it? Resources? Stability? Complexity??
  
  2. the cancellation in memoryis TRUE. What is your recommendation? Always use? Never use? Always use on OLTP? Never use on BI?
  
  Thank you!

  I'm not an expert, but still, just a few thoughts,.

  How much memory is consumed for a database by using this feature? Are there other 'costs' associated with it? Resources? Stability? Complexity??

  IMU is maintained by the pools to cancel it. If I remember correctly, the IMU pools are maintained by an undocumented parameter and is stored in the shared pool. The 'cost' so that if you plan to upgrade to 10g (it's one of 10g featured), it would take an additional memory that you allocate to the pool shared from your previous version. According to oracle, the additional memory overhead would be about 30 to 35%, which must be added. In addition, it is not always that oracle would use IMU. If the pools are filled, the traditional mechanism to cancel (and thus restore since its also kept in the shared pool as redo private) is used. In addition, to use a lock lock IMU is used and too much use that can show you an IMU lock contention in your statspack/awr report. This, if you want to talk, can be attributed as a cool resource overhead for using this feature. I'm not sure that it can be called complex because you're supposed to not 'play' with thisfeature, sound here and his work, that's all! I have not met many people who knew even before I tell them that this feature still exists! So, if you do not break what is already working, I guess that things would be fine.

  The default value for memoryundo is TRUE. What is your recommendation? Always use? Never use? Always use on OLTP? Never use on BI?

  Even what I said just above, this is a feature that is there and is working out-of-the-box. Its designed to work and work. For a BI/DSS environment, I guess that's not a lot of use (no use would also be reason to say since there was almost no DML) and for OLTP, heavily used. As I said, I'm not an expert so wait for gurus like Daniel, Jonathan, Billy, Charles, Hemant (the list is too long) to submit their comments thereon.

  HTH
  Aman...

• Database of slowdown

  Dear friends,
  
  The developers around me tell me that the database is slow (they use PL/SQL Developer and other tools).
  
  How can I check if the database is actually causing the problem?
  
  Thanks in advance.
  
  Nith

  I know not why CKPT said of the response time of the listener since once the connection is estabilished, listener is no longer in the picture, but probably it's guessing all the time to take the implementation of connection as well as why he believes it. Anyway, there is no way toget it as long as you enable tracing for the listener. Who should inquire about the activity of the listener. But I would rather check not it but would take a statspack/awr report and will see what comes out.

  HTH
  Aman...

• How to put the code in format fixed

  Hello
  

  The post of code, report statspack/AWR, plans of execution or trace files, start and end of the article with the tag

   >(lowercase, curly brackets, no spaces) so that the text appears in
  fixed format

  If I want to paste the results to txt file or sql * plus which should appear in a fixed format here for the understanding of people, so how can I do this please.
  
  I tried to put {at the beginning and} at the end and checked by clicking on the preview, but again its all messed up tab.
  
  Please help with some examples so that I can continue with it.
  
  Thank you.
  
  Concerning
  KAM

  Please refer to http://forums.oracle.com/forums/help.jspa

  When you type:

  {noformat}

  select *
  from   dual;
  

  {noformat}

  It appears that:

  select *
  from   dual;
  

  When you post the it.

• Consolidation of DB - Cost Cutting

  Dear members,
  I'm new to this forum. We have several servers classified as development, Test and Production in which several databases are executed in each of them. Can anyone suggest me how to proceed with the consolidation of oracle database in the server group. I have an idea little lit up like
  patterns in several databases can be grouped into a single oracle database.
  
  But how calculate efficiency as, a particular database have not been used for the last 2 months... etc so that it can be used as a factor of consolidation.
  
  If I'm not mistaken, the size of each scheme can be determined in dba_segments selection owner and sum column (bytes).
  
  Please let me know what might be other issues to look out here in the consolidation.
  
  Kind regards
  Mani
  TCS

  First of all, you must understand what databases are candidates for consolidation. For example
  -Your databases do support packaged applications with specific requirements for the version parameters and/or database initialization? If so, these databases must usually be retained separately.
  -L' one of your databases have other reasons to run a different Oracle that everyone else version (additional construction or different and cycles, different points on the life cycle, etc.)?
  -Your data have the same availability requirements? If not, it can get very difficult (and therefore expensive) to coordinate downtime if you consolidate the databases that have needs different downtime.
  -What's your approach to the consolidation of the support of the permissions? If you regularly give users (particularly in environments below) powerful privileges (i.e. SELECT ANY TABLE) knowing that the data in this particular database is a relatively low risk, consolidate with a database that contains payroll information is probably a bad idea.
  -Your applications use public synonyms? If so, these synonyms would be in line with consolidation?

  Second, what is the efficiency as you try to win? If a particular database has not been used in 2 months, why is - it? Assuming it will be reused, really saving you much if you consolidate with another database on the same server? You can decrease the RAM required on the server, but it is often the least expensive component to add. If you can consolidate enough bases to start to get rid of servers, which can begin to be a reduction of costs (especially where you have CPU based licenses).

  Assuming that you have a level of oversight in place, either at the level of the OS or Oracle via statspack / AWR, you can usually get an idea of the average and which are placed on a system to see what you can gain by consolidating the peak loads. But you need determine how correlated is the use of various databases - if takes the data base to 100% of the CPU and database B takes 100% of CPU, you might be well consolidate if one runs almost exclusively during the night and B works almost exclusively during the day. On the other hand, if you have databases with vertices coincide (i.e. the activity in a single system implies that it is likely to be activity in another), you can get very little consolidation and can end up creating new bottlenecks if everyone is fighting on the CPU during a peak. And you will need to know how much you should allocate slack (for example, some applications may spec on the server of database planning for a few years worth of growth - you will need to have access to these projections to determine whether the database that only takes 10% of the CPU will double or triple in size in the next few years).

  Justin

• statspack and awr is to collect statistics, or simply to present

  Hi all
  If I have a db with statspack installed in 9i
  I took a glance at 13:00 then took antother wink at 15:00
  
  so now I can analyze the period between 13:00 and 15:00
  
  My question is, if statspack is NOT installed, is the db should run FASTER, because nothing is the collection of statistics
  the statistics are collected in any case or statspack is just a way to get this perspective
  
  and what about the AWR in 10g?
  
  I have red this link
  http://download.Oracle.com/docs/CD/B10501_01/server.920/a96533/statspac.htm#32191
  but it do not answer this point.
  
  Thank you very much

  STATSPACK and AWR do not run continuously and collect statistics, so that they cause no impact on performance.

  When you run the blink of an eye it just reads the current statistics of the system to the $ V and X tables $ and stores them in permanent tables. The V$ X$ tables are always up-to-date even if statspack is not installed.

  Published by: Robert Geier on October 13, 2009 13:06

• How can I save an AWR/statspack report in xml format?

  How can I save an AWR/statspack report in xml format? or is there a way... coz when you create the report, it offers the possibility for normal (the base sql or txt file) or html... of 10.2.0.3

  Hello

  H4. No format html and txt. will not be properly charged to the top, you will need to extract using the suite to generate .dmp

  This is how to extract the AWR data and

  sqlplus "/as sysdba"
  SQL> @$ORACLE_HOME/rdbms/admin/awrextr.sql
  

  and that's how to load upwards

  sqlplus "/as sysdba"
  SQL>@$ORACLE_HOME/rdbms/admin/awrload.sql
  

  Concerning

• AWR with Oracle 11g Standard edition

  Hello

  I have the database to Oracle 11 g Release 11.2.0.1.0 - 64 bit Production in my database. And when I ask for the control_management_pack_access parameter it value as DIAGNOSTIC + TUNING. I generated several AWRs with 1 hour 24 hours period as well. The two reports gives the error WARNING: as DB time is less than a second, he had minimal activity of foreground in period snapshot. Some of the percentage values are not valid. and DB time is zero for each AWR reports.

  Can someone please help me solve this confusion DIAGNOSTIC + TUNING pack comes with the Standard edition and why DB time gives zero?

  Thank you

  TK.

  Hello

  Database created on Standard Edition should have control_pack_management_access = none. You can change, but the risk is that you use the options that are available only on the Enterprise Edition. In itself, you can use Statspack.

  Kind regards

  Franck.

• using AWR se edition

  Hello

  the 11.2.0.4 SE One on Win 2008

  SE edition license includes the use of AWR?

  A I see in the documentation Automatic workload management is not included. But is it the same as CWA (automatic workload repository)?

  
  

  Thank you.

  user10274093 wrote:

  Hello

  the 11.2.0.4 SE One on Win 2008

  SE edition license includes the use of AWR?

  A I can see in the documentation that automatic management of the workload is not included. But is it the same as CWA (automatic workload repository)?

  Thank you.

  AWR is only supposed tobe used if you have access to the Diagnostics and Tuning pack license. So if you do not, your only option is to use Statspack.

  Aman...

• Statistics in AWR

  Hi all

  11.2.0.3

  AIX 6.1

  I run stats AWR gathering every hour on top of the hour, in our PRODUCTION BOM of database 24 hours a day. Is it possible to select the window (that stretch of time) in a certain day was our most frequent database, which means more intensive resources? just by using a sql query? or I have to run each time slot 24 times and compare 24 report results?

  Thank you very much

  zxy

  Hello

  Looks like you're confused about basic AWR. Given AWR do not live in a special database. It is stored in the same database in tablespace SYSAUX, SYS. Tables of $ WRH which are exposed via SYS. % DBA_HIST views built on top of them. That's what I use in my example above.

  If you do not have instant AWR in your database, then you have not all historical data about the performance of your system. Then you need to either activate snapshot AWR collection or (if you don't have the Diagnostics pack license or do not want to use AWR for other reasons) use statspack.

  Best regards

  Nikolai

• CPU without AWR

  Hello
  on the 11g on Win 2008 R2.
  We cannot use AWR report.
  Windows System Administrator said that, on 23/04/13 22:39, oracle.exe process used 62% of CPU.
  
  I wanted to know what was my database. So I ran the following:

  SQL> select count(*) from V$SQLAREA where last_load_time
    2  BETWEEN
    3  to_date('23.04.2013 22:00:00', 'dd.mm.yyyy HH24:MI:SS')
    4  AND to_date('23.04.2013 23:00:00', 'dd.mm.yyyy HH24:MI:SS')
    5  ;
  
    COUNT(*)
  ----------
           0

  And nothing special in alertlog.
  This means that this database did nothing? Where otherwise I should look to see what was going on?
  
  Thanks for help.

  Hello

  You have no statspack report?
  Kind regards

• AWR and report STATPACK

  Hello
  on the 11g R2,.
  1 are necessary the AWR and STATPACK reports?
  
  2 AWR isn't enough, we have all by itself?
  
  Thank you.

  1 are necessary the AWR and STATPACK reports?

  No, only AWR enough with the latest features

  2 AWR isn't enough, we have all by itself?

  Yes, however statspack is free, but the need with the AWR additional license.

• AWR

  Hi all
  
  How will we change the AWR snap_level. By default, it displays as 1.

  SQL> select distinct snap_level from dba_hist_snapshot;
  
  SNAP_LEVEL
  ----------
        1

  In statspack, we can change the snap_level

  user13355115 wrote:
  But even after changing the statistics_level = all the snap_level is always 1.

  My question is as statspack can we change and increase the snap_level?

  exec dbms_workload_repository.create_snapshot(flush_level => 'ALL');
  

  It will give you Snap_level = 2

• process analasys AWR

  Hi s/n,.
  How to analyze the AWR reports, what are the important areas there is no limit here.
  
  Thank you
  Tmadugula

  AWR and Statspack are similar so this series should help: http://jonathanlewis.wordpress.com/statspack-examples/

  (at present two documents from Oracle at the top of the list have been moved (again!) but I'll try to find out where the next week.

  Concerning
  Jonathan Lewis