Performance check after CPU

Dear colleagues,
Is any posilble check or test (benchmark) of DB performance before and after CPU patching?
Any testing SAP or SQL report, which makes a load on DB?
Thank you for advice.
Best Regards
Jakub Vaněk

> Is any posilble check or test (benchmark) of DB performance before and after CPU patching?
> Any testing SAP or SQL report, which makes a load on DB?
Those test would be synthetic and may or may not reflect your current environment.
I'd check some long running reports in your environment if you see any difference.
Markus

Similar Messages

Perform check after release step of purchasing requisition

Hi,
does anybody have a suggestion where to place best some logic which needs to be performed after a release step is performed successfully.
background: we use different types of Purchase requisitions, while each one is assigned to a static release strategy. I like to change it in more dynamic way by adding certain logic after a release step is performed.
example: if preceeding release step was 10 and purchase requisition is of type x then check purchase value and account assignment in order to process the current stepo automatically.
Currently the logic runs during a nightly job, but I would like to add it immediately after a release has been performed. I would imagin something like user exit, BusinessAddIn, or an event which I can use assign the job.
Can anybody suggest something?
Thanks in advance
Best regards
Torsten
Edited by: Torsten Hübsch on Apr 29, 2011 10:36 PM

Hi,
Enhancement M06B0005 is relevant for PR in case of overall release
Check on trx SMOD :
Enhancement        M06B0005
Short text         Changes to comm. structure for overall release of requisistion
--> Function module                Short Text
EXIT_SAPLEBND_004              Changes to Communication Structure for Over.
Enhancement        M06B0002 is for item-release
Short text         Changes to comm. structure for purchase requisition release
-->
Function module
EXIT_SAPLEBND_001
BR
Nadia Orlandi

Performance drop after CPU upgrade.

Hias,
I got a new CPU.
I expected some performace gain icw the old one.
However...it is a lot slower
I went from 64.000 to 47.000 in aquamark. Didn't record the values axeactly, but it seems like the CPU part is the one that is lagging behand.
I did need to reinstall and upgrade vid.card drivers and sound card drivers and CPU drivers.
Any1 have a clue the cause of this.

Sorry, should have it a bit more clear:
http://valid.x86-secret.com/show_oc?id=44047 <-old CPU (3200+ Winnie)
http://valid.x86-secret.com/show_oc?id=68314 <-new CPU (Opeteron 148)
I upgraded bios from 3.3 to 9.2.
I didn't change any other HW, As u can see (follow the links) memory settings are about the same.

SQL performance drops after 20 mins

Hello,
I am using Oracle Version 9.2.0.5
I have an SQL query that starts off fast but the performance drops after running for 20 mins. It continues to drop till a point where it becomes painfully slow.
The query is as attached below:
SELECT      /*+ INDEX (abe ASR_I_BID_PRD_TOD_5FAX)
INDEX (aae ASR_I_AE_RC_BR)
INDEX (ab ASR_UK_BASES)
INDEX (aaac ASR_I_AFAC_SP)
+*/aae.f8_code
||','||     aae.sector
||','||     aae.ouc
||','||     aae.rule_type
||','||     aae.base_ref
||','||     aae.prodlist_ref
||','|| aaac.L101
||','|| aaac.L201
||','|| aaac.L301
||','|| aaac.L401
||','|| aaac.L501
||','|| aaac.L601
||','|| aaac.L701
||','|| aaac.L801
||','|| aaac.L802
||','|| aaac.L830
||','|| aaac.L850
||','|| aaac.L870
||','|| aaac.L901
||','|| aaac.L1001
||','||     LTRIM(TO_CHAR((aae.disaggregated_amount*abe.percentage*aaac.chain_perc)/10000,'999999999999999.99'))
FROM asr_owner.asr_account_entries aae
, asr_owner.asr_base_elements_5fax abe
, asr_ff_audit_chains aaac
, asr_owner.asr_bases ab
WHERE aae.run_code = '5fax'
AND ab.run_code = aae.run_code
AND ab.base_type = aaac.base_type
AND ab.base_id = abe.base_id
AND ab.base_ref = aae.base_ref
AND aaac.base_type = 'L'
AND aaac.business <> 'R'
AND aae.trans_type IN ('A','C','D')
AND abe.base_id = aae.base_id
AND 100*(abe.exhaust_level)+abe.exhaust_sub_level = 101
AND aaac.start_prod = abe.product
AND ABS((aae.disaggregated_amount*abe.percentage*aaac.chain_perc)/10000) >= 0.01
It has huge number of records. I have checked all join conditions and they are all fine.
asr_owner.asr_account_entries - 3.6 millions rows
asr_owner.asr_base_elements_5fax - 1.8 million rows
asr_ff_audit_chains - 5.8 million rows
asr_owner.asr_bases - 0.5 million rows
The autotrace and plan are as follows
Statistics
0 recursive calls
0 db block gets
5264 consistent gets
3522 physical reads
0 redo size
525 bytes sent via SQL*Net to client
456 bytes received via SQL*Net from client
1 SQL*Net roundtrips to/from client
5 sorts (memory)
0 sorts (disk)
0 rows processed
PLAN_TABLE_OUTPUT
| Id | Operation | Name | Rows | Bytes | Cost |
| 0 | SELECT STATEMENT | | 83 | 13529 | 12465 |
| 1 | TABLE ACCESS BY INDEX ROWID | ASR_FF_AUDIT_CHAINS | 83 | 6225 | 617 |
| 2 | NESTED LOOPS | | 83 | 13529 | 12465 |
| 3 | NESTED LOOPS | | 1 | 88 | 11848 |
| 4 | NESTED LOOPS | | 2201 | 113K| 3044 |
| 5 | TABLE ACCESS BY INDEX ROWID| ASR_BASES | 774 | 13158 | 722 |
| 6 | INDEX FULL SCAN | ASR_UK_BASES | 104K| | 104 |
| 7 | TABLE ACCESS BY INDEX ROWID| ASR_ACCOUNT_ENTRIES | 3 | 108 | 3 |
| 8 | INDEX RANGE SCAN | ASR_I_AE_RC_BR | 8 | | 1 |
| 9 | TABLE ACCESS BY INDEX ROWID | ASR_BASE_ELEMENTS_5FAX | 1 | 35 | 4 |
| 10 | INDEX RANGE SCAN | ASR_I_BID_PRD_TOD_5FAX | 1 | | 3 |
| 11 | INDEX RANGE SCAN | ASR_I_AFAC_SP | 8241 | | 9 |
Note: cpu costing is off, PLAN_TABLE' is old version
19 rows selected.
I have also tried without the hints and the plan looks like
| Id | Operation | Name | Rows | Bytes | Cost |
| 0 | SELECT STATEMENT | | 83 | 13529 | 679 |
| 1 | TABLE ACCESS BY INDEX ROWID| ASR_FF_AUDIT_CHAINS | 83 | 6225 | 617 |
| 2 | NESTED LOOPS | | 83 | 13529 | 679 |
| 3 | HASH JOIN | | 1 | 88 | 370 |
| 4 | HASH JOIN | | 2201 | 113K| 208 |
| 5 | TABLE ACCESS FULL | ASR_BASES | 774 | 13158 | 16 |
| 6 | TABLE ACCESS FULL | ASR_ACCOUNT_ENTRIES | 12125 | 426K| 191 |
| 7 | TABLE ACCESS FULL | ASR_BASE_ELEMENTS_5FAX | 18325 | 626K| 161 |
| 8 | INDEX RANGE SCAN | ASR_I_AFAC_SP | 8241 | | 9 |
Note: cpu costing is off, PLAN_TABLE' is old version
16 rows selected.
I have tried increasing the hash_area_size parameter to 20MB (it is currently 10MB).
I have tried setting the tuning parameters like optimizer_index_cost_adj but to no success.
Could you please help.

Thanks a lot for all the help.
Here is another query that is running quite slow. Any help on this might be helpful
EXPLAIN PLAN SET statement_id = 'auditextract' FOR
SELECT /*+NO_INDEX(aaac)
       NO_INDEX(aae)
       NO_INDEX(abe)
       NO_INDEX(ab)
       NO_INDEX(agp)+*/
     aae.run_code
,     aae.f8_code
,     aae.sector
,     aae.ouc
,     aae.rule_type
,     aae.base_ref
,     aae.prodlist_ref
,     aae.trans_type
,     aaac.base_type
,DECODE(NVL(aaac.exhaust_chain,'NULL'),'NULL',',',NULL)||aaac.exhaust_chain||
DECODE(NVL(aaac.exhaust_chain,'NULL'),'NULL',NULL,',')||LTRIM(RTRIM(aaac.final_prod))||
            RPAD(',',14-DECODE(aaac.final_level,1001,14,901,13,870,12,850,11,830,10,802,9,801,8,701,7,601,6,501,5,401,4,301,3,201,2,101,1),',') exhaust_chain,
     aaac.final_level     ,
     aaac.business ,
     L101,
     L101_TOD,
     L201,
     L201_TOD,
     L301,
     L301_TOD,
     L401,
     L401_TOD,
     L501,
     L501_TOD,
     L601,
     L601_TOD,
     L701,
     L701_TOD,
     L801,
     L801_TOD,
     L802,
     L802_TOD,
     L830,
     L830_TOD,
     L850,
     L850_TOD,
     L870,
     L870_TOD,
     L901,
     L901_TOD,
     L1001,
     L1001_TOD,
     (aae.disaggregated_amount*abe.percentage*aaac.chain_perc)/10000 money
FROM   asr_owner.asr_account_entries aae
,      asr_owner.asr_base_elements_5fax abe
,      asr_owner.asr_bases ab
,      ASR_FF_AUDIT_CHAINS aaac
WHERE aae.run_code    = '5fax'
AND    ab.run_code = aae.run_code
AND    abe.run_code = ab.run_code
AND    ab.base_type = aaac.base_type
AND    ab.base_id = abe.base_id
AND    aaac.base_type = 'L'
AND    aaac.business = 'R'
AND    ab.base_ref = aae.base_ref
AND    aae.trans_type IN ('A','C','D','G','J')
AND    abe.base_id     = aae.base_id
AND    abe.exhaust_level = 1
AND    abe.exhaust_sub_level = 1
AND    aaac.start_prod = abe.product
AND    NOT EXISTS (SELECT 1
                    FROM asr_group_prods agp
                    WHERE agp.product_group IN ('SW500','SR500','G176')
                    AND (aaac.L101 = agp.product OR
                              aaac.L201 = agp.product OR
                                 aaac.L301 = agp.product OR
                                 aaac.L401 = agp.product OR
                                 aaac.L501 = agp.product OR
                                 aaac.L601 = agp.product OR
                                 aaac.L701 = agp.product OR
                                 aaac.L801 = agp.product OR
                                 aaac.L802 = agp.product OR
                                 aaac.L830 = agp.product OR
                                 aaac.L850 = agp.product OR
                                 aaac.L870 = agp.product OR
                                 aaac.L901 = agp.product OR
                                 aaac.L1001 = agp.product))
AND    abs((aae.disaggregated_amount*abe.percentage*aaac.chain_perc)) >= 100
PLAN_TABLE_OUTPUT
| Id | Operation             | Name                   | Rows | Bytes | Cost |
|   0 | SELECT STATEMENT      |                         |   571 |   112K| 5586 |
|* 1 | FILTER               |                         |       |       |       |
|* 2 |   HASH JOIN           |                         |   571 |   112K| 5586 |
|* 3 |    HASH JOIN          |                         |    53 | 5406 |   370 |
|* 4 |     HASH JOIN         |                         | 2974 |   177K|   178 |
|* 5 |      TABLE ACCESS FULL| ASR_BASES               |   770 | 15400 |    16 |
|* 6 |      TABLE ACCESS FULL| ASR_BASE_ELEMENTS_5FAX | 26178 | 1048K|   161 |
|* 7 |     TABLE ACCESS FULL | ASR_ACCOUNT_ENTRIES     | 17739 |   710K|   191 |
|* 8 |    TABLE ACCESS FULL | ASR_FF_AUDIT_CHAINS     |   126K|    12M| 5215 |
|* 9 |   TABLE ACCESS FULL   | ASR_GROUP_PRODS         |     1 |    10 |     2 |
Predicate Information (identified by operation id):
   1 - filter( NOT EXISTS (SELECT /*+ */ 0 FROM "ASR_GROUP_PRODS" "AGP" WHERE
              ("AGP"."PRODUCT_GROUP"='G176' OR "AGP"."PRODUCT_GROUP"='SR500' OR
              "AGP"."PRODUCT_GROUP"='SW500') AND ("AGP"."PRODUCT"=:B1 OR "AGP"."PRODUCT"=:B2
              OR "AGP"."PRODUCT"=:B3 OR "AGP"."PRODUCT"=:B4 OR "AGP"."PRODUCT"=:B5 OR
              "AGP"."PRODUCT"=:B6 OR "AGP"."PRODUCT"=:B7 OR "AGP"."PRODUCT"=:B8 OR
              "AGP"."PRODUCT"=:B9 OR "AGP"."PRODUCT"=:B10 OR "AGP"."PRODUCT"=:B11 OR
              "AGP"."PRODUCT"=:B12 OR "AGP"."PRODUCT"=:B13 OR "AGP"."PRODUCT"=:B14)))
   2 - access("SYS_ALIAS_14"."START_PROD"="ABE"."PRODUCT")
       filter(ABS("AAE"."DISAGGREGATED_AMOUNT"*"ABE"."PERCENTAGE"*"SYS_ALIAS_14".
              "CHAIN_PERC")>=100)
   3 - access("AB"."BASE_REF"="AAE"."BASE_REF" AND
              "ABE"."BASE_ID"="AAE"."BASE_ID")
   4 - access("AB"."BASE_ID"="ABE"."BASE_ID")
   5 - filter("AB"."RUN_CODE"='5fax' AND "AB"."BASE_TYPE"='L')
   6 - filter("ABE"."RUN_CODE"='5fax' AND "ABE"."EXHAUST_LEVEL"=1 AND
              "ABE"."EXHAUST_SUB_LEVEL"=1)
   7 - filter("AAE"."RUN_CODE"='5fax' AND ("AAE"."TRANS_TYPE"='A' OR
              "AAE"."TRANS_TYPE"='C' OR "AAE"."TRANS_TYPE"='D' OR "AAE"."TRANS_TYPE"='G' OR
              "AAE"."TRANS_TYPE"='J'))
   8 - filter("SYS_ALIAS_14"."BASE_TYPE"='L' AND "SYS_ALIAS_14"."BUSINESS"='R')
   9 - filter(("AGP"."PRODUCT_GROUP"='G176' OR "AGP"."PRODUCT_GROUP"='SR500' OR
              "AGP"."PRODUCT_GROUP"='SW500') AND ("AGP"."PRODUCT"=:B1 OR "AGP"."PRODUCT"=:B2
              OR "AGP"."PRODUCT"=:B3 OR "AGP"."PRODUCT"=:B4 OR "AGP"."PRODUCT"=:B5 OR
              "AGP"."PRODUCT"=:B6 OR "AGP"."PRODUCT"=:B7 OR "AGP"."PRODUCT"=:B8 OR
              "AGP"."PRODUCT"=:B9 OR "AGP"."PRODUCT"=:B10 OR "AGP"."PRODUCT"=:B11 OR
              "AGP"."PRODUCT"=:B12 OR "AGP"."PRODUCT"=:B13 OR "AGP"."PRODUCT"=:B14))
Note: cpu costing is off
47 rows selected.

Performance issue after Upgrade from 4.7 to ECC 6.0 with a select query

Hi All,
There is a Performance issue after Upgrade from 4.7 to ECC 6.0 with a select query in a report painter.
This query is working fine when executed in 4.7 system where as it is running for more time in ECC6.0.
Select query is on the table COSP.
SELECT (FIELD_LIST)
        INTO CORRESPONDING FIELDS OF TABLE I_COSP PACKAGE SIZE 1000
        FROM COSP CLIENT SPECIFIED
        WHERE GJAHR IN SELR_GJAHR
          AND KSTAR IN SELR_KSTAR
          AND LEDNR EQ '00'
          AND OBJNR IN SELR_OBJNR
          AND PERBL IN SELR_PERBL
          AND VERSN IN SELR_VERSN
          AND WRTTP IN SELR_WRTTP
          AND MANDT IN MANDTTAB
        GROUP BY (GROUP_LIST).
   LOOP AT I_COSP      .
     COSP                           = I_COSP      .
     PERFORM PCOSP       USING I_COSP-_COUNTER.
     CLEAR: $RWTAB, COSP                          .
     CLEAR CCR1S                         .
   ENDLOOP.
ENDSELECT.
I have checked with the table indexes, they were same as in 4.7 system.
What can be the reson for the difference in execution time. How can this be reduced without adjusting the select query.
Thanks in advance for the responses.
Regards,
Dedeepya.

Hi,
ohhhhh....... lots of problems in select query......this is not the way you should write it.
Some generic comments:
1. never use SELECT
                   endselect.
   SELECT
into table
   for all entries in table
where.
   use perform statment after this selection.
2. Do not use into corresponding fields. use exact structure type.
3. use proper sequence of fields in the where condition so that it helps table go according to indexes.
    e.g in your case
          sequence should be
LEDNR
OBJNR
GJAHR
WRTTP
VERSN
KSTAR
HRKFT
VRGNG
VBUND
PARGB
BEKNZ
TWAER
PERBL
sequence should be same as defined in table.
Always keep select query as simple as possible and perform all other calculations etc. afterwords.
I hope it helps.
Regards,
Pranaya

Check average CPU utilization in AIX 5.3

Hi All,
could anyone please share knowledge on the following:
How to check average CPU utilization in AIX 5.3.What command is used in AIX to check this output?
Would appreciate an early reply!
Thanks for your time!
Regards,

curt: Reports CPU utilization for each kernel thread (beginning with AIX 5.2)
iostat: Displays utilization data for (Terminals, CPU, Disks, Adapters (beginning with AIX 5.1))
netpmon: Uses the trace facility to report on network activity, including CPU consumption
pprof: Reports CPU usage of all kernel threads over a period of time
ps: Displays statistics and status information about the processes in the system, such as CPU utilization
time, timex: Prints the elapsed and CPU time used by the execution of a command
vmstat: Displays VMM data, such as CPU utilization
[Performance reporting and analysis commands |http://publib.boulder.ibm.com/infocenter/systems/index.jsp?topic=/com.ibm.aix.prftungd/doc/prftungd/perf_report_anlysis_comms.htm]
You can also use:
TPROF
http://www.ibm.com/developerworks/systems/articles/AIX5.2_performancetoolsupdatepart3.html

Performing check and potential update for status control table

Cheers Experts,
I searched in the forum, but I didn't find an answer.
I'm working on a 3.0B BW system.
After updating from PSA to an ODS, in the log I see the following line:
"Performing check and potential update for status control table" and double clicking on it, it says:
If data is loaded into an InfoCube, or existing data is edited (aggregated/compressed/deleted/got from a DataMart), then there is a change in the potential reportability of the data , or the possibility of deleting data by request, or of aggregating or compressing.
This status of each one of these Cubes is stored in a status table, that is updated when there is any change to the status of a request in the Cube.
The system now analyzes the requests in the Cube and compares the calculated status with the status table.
If deviations from the status table arise then you are given the option of adjusting the status table.
I have 2 questions:
1) What is BW doing, mainly because I'm working on an ODS and not an Infocube?
2) What can I do to reduce its execution time, since it lasts several minutes?
Thanks and greetings.
Roberto

Roberto, did you ever find an answer to your question? We also are experiencing long runtimes during this particular process running 3.x deltas from ODS to ODS and cubes.

Checklist to check after SP upgrade

Hi Experts,
We are planning a system upgrade from SP15 to SP27 very soon.
Is there any kind of check-list which can provide us the options to check after post-upgrade of the system?
Any kind of help/ documents/links related with checklist would be helpfull.
Note that the current version being used is BI 7.0
Regards,
Arpit
Edited by: apkmnp on Mar 8, 2012 11:58 AM
Edited by: apkmnp on Mar 8, 2012 11:59 AM

Hi,
The basic checklist for a SP upgrade would contain tasks like below
Before Upgrade:
-check if the objects are active.
-initialize the deltas
After Upgrade:
-create a infopackage/dso/cube etc
-run the delta DTPs
-run the full DTPs
-run the delta InfoPackage
-run the full InfoPackage
-run the APDs
-execute the queries in RSRT
-execute the process chain
-edit the psa
-delete data from PSA, DSO, changelog, cube
-load data to the dso cube etc
-comress your cube
-Extract data from R3/ XI (PI) if connected or any other source system.
-load your masters
-run the attribute change run for the masters
-run your important programs
-post data in R3 and pull it to BW
-Create workbook
-create query variants
And if any critical or specific tasks you perform you would have to check those.
Hope this helps
Regards,
Joe

Performance check for status control table

Hi,
When ever we activate any ODS or delete the request ,a message 'Performance Check for Status control Table ' appears.
What does it mean?Please explain what is status control table and how it performs?
Thanks & Regards,
Dinakar.

Hi,
This message is not a issue in itself. It is just a SAP message which means that the system is checking/updating the system tables to carry out the operation whether it can be done or not.
If your loads are failing at this step or after this message there can be many reasons, check some of the possible reasons below
Performing check and potential update for status control table
Update from PSA error in Process Chain
problem in deleting request
Reporting not available ,even if Data is sucessfully loaded into DT
Hope this helps,
Kush kashyap

Performance issue after DB copy

Hello,
I am facing performance issue after doing database copy. Let me explain entire story.
Enviornment is Oracle 10g (10.2.0.4), Operating System is Solaries.
We have one production system with DB sid (PR1) and i have to builded new system using PR1 as PR2, we have done full offline copy from one hardware to another hardware (Hardware is same on PR1 and PR2) and perform rsync command at os level to sync all oracle files at OS level(PR2 is not build using backup/restore or export import method).
Now problem is that
One of query if i am running on PR1 system it is taking only 10-15 seconds, while if i am running same query on PR2 it is taking nearly about 55-60 mins. In initial search i found SGA is very diffrent from PR1 to PR2 then i did change the SGA parameter like PR1 in PR2 after doing changes i am running same query it is taking 6-7 mins, so i got some benifit here, but still i need to tune the query. Query contains 3 tables and all tables having atleast 3-4 indexes.
Can some one help what to tune and what to check?
Thanks,
Singh

Check the explain plan on both systems to see if both instances are trying to do the same thing. This will tell you a lot about what is going on based on physical or logical reads and how long they are taking. You should have the same statistics on both databases if you did a physical copy, but check that too just to make sure. If stats and explain plan are the same, check your hardware. Do you have the same amount of RAM on both systems? Are you using similar storage? If the plan is showing that you are doing a considerable amount of physical I/Os on both it could just be that your physical disk is slower on server two.

System performance degrades after server migration ???

Hi Friends,
System performance degrades after we migrate our BW 3.5 Server (Production) from UK to Germany.
for details :
1. Data is coming to Informatica server from Informatica its going to POSDM Server (Point of Sales Data Management) from POSDM we are running pipes and data is coming to BW (Delta Queqe)
so before Server Migration it was taking 2 Hrs to load 4 Million Records
After Server Migration it is taking 4 Hrs.
Please help out to find the reason for this.
Note : Server Ram , Hard Disk , Speed is same on both Servers
Thanks
Asim

Note : Server Ram , Hard Disk , Speed is same on both Servers
Are you cahnge any application or database parameter (the OS and all patch'es are the same?)
How you doing migration? Very low information... Are you check the network configuration are the same( for example check the network speed 1G or 100MB).
Are you trying to analyse the St03N and St04 t-codes?
Regards.

Performance degradation after setting filesystemio_option=setall from none.

Hi All,
We have facing performance degradation after setting filesystemio_option=setall from none on my two servers as mentioned below.
Red Hat Enterprise Linux AS release 4 (Nahant Update 7) 2.6.9 55.ELhugemem (32-bit)
Red Hat Enterprise Linux Server release 5.2 (Tikanga) 2.6.18 92.1.10.el5 (64-bit)
We are seeing lots of Disk I/O happening. We expected "*filesystemio_option=setall* " will improve performance but it is degrading. We getting slowness complains.
Please let me know do we need to set somethign else along with this ...like any otimizer parameter( e.g. optimizer_index_cost_adj, optimizer_index_caching).
Please help.

Hi Suraj,
<speculation>
You switched filesystemio_options to setall from none, so, the most likely reason for performance degradation after switching to setall is the implementation of directio. Direct I/O will skip the filesystem buffer cache, and and allow Oracle to read directly from disk to the database buffer cache. However, on a system where direct I/O is not implemented, which is what you had until you recently messed with that parameter, it's likely that you had an undersized database buffer cache, but that was ok, because many (most) of the physical I/Os your database was doing, were actually being serviced by the O/S filesystem buffer cache. But, you introduced direct I/O, and wiped out the ability of the O/S to service any physical I/Os from filesystem buffer cache. This means that every cache miss on the database buffer cache, turns into a real, physical, spin-the-disk, move-the-drive-head, physical I/O. And, you are suffering the performance consequences.
</speculation>
Ok, end of speculation. Now, assuming that what I've outlined above is actually going on, what to do? Why is direct I/O lower performing than buffered, non-direct I/O? Shouldn't it's performance be superior?
Well, when you have an established system that's using buffered I/O, and you switch to direct I/O, you almost always will have to increase the size of the database buffer cache. The problem is that you took a huge chunk of memory away from the the O/S, that it was using to buffer your I/Os and avoid physical I/O. So, now, you need to make up for it, by increasing the size of the database buffer cache. You can do this, without buying more memory for the box, because the O/S is no longer going to need to use so much memory for filesystem buffers.
So, what to do? Is it worth switching? Well, on balance, it makes sense to use direct I/O, and give Oracle a larger database buffer cache, for the simple fact that (particularly on a server that's dedicated to being an Oracle database server), Oracle has far more sophisticated caching algorithms, and a better understanding of the various types of data being cached, and so should be able to make more efficient use of the memory, than the (relatively) brain dead caching algorithms of the kernel and filesystem mechanisms.
But, once again, it all comes down to this:
What problem are you trying to solve? Did you have any I/O related issues? Do you have any compelling reason to implement direct I/O? Rule #1 is "if it ain't broke, don't fix it." Did you just violate rule #1? :-)
Finally, since you're on Linux, you can use the 'free' command to see how much memory is on the box, how much is free, and how much is dedicated to filesystem cache buffers. This response is already pretty long, so, I'm not going to get into details, however, if you're not familiar with the command, the results could be misleading. Read the man page, and try to be clear about understanding it before you make any assumptions about the output.
Hope that helps,
-Mark

Performance Hit After Oracle Database Upgrade to 10.2.0.4

We have a couple dozen workbooks that took this performance hit after the upgrade of the database/migration to a new server. Worksheets that executed in the ten second range are now running for hours or simply not finishing. We took the new server factor out of the equation by rolling back the database to 10.2.0.3 where a test EUL resides and the problem was resolved. Has anyone seen this issue? Does anyone have an suggestions? An early reply would be greatly appreciated.
Thanks,
Jerre

Rod,
Thanks for the quick reply. We are looking at the different plans and modifying the optimizer settings, switching back and forth, as we speak. We are now starting with the hints. Currently our Server 'optimizer_mode' parameter is ALL_ROWS. We are planning to change the to 'Choose' and see what happens. The workbooks that are impacted are on our oldest business areas of Finance and HR. The former setup was borrowed from another school for a quick, low cost start up. The latter was thrown together by novices. Our true datamarts developed by knowledgeable personnel with star schemas are not impacted. Of course we are planning on redoing the older business areas but time, personnel and money matters slow things down. It is these workbooks on the older business areas that are greatly affected by the migrations and upgrade. We eventually get things to settle down but past actions do not always have the same resolution with newer and better servers and upgrades.
Thanks,
Jerre

What are the things to check after full imp in 10g?

Dear all,
Source
=======
OS server ==> HP-UX
Oracle version ==> Oracle 9.2.0.8
DB Name ==> MSST
DB total users ==> 320
Full export in ==> /u02/export/Jan09.dmp
Tablespaces ==> SC, SC_I, SA, SA_I, PP, PP_I, AC, AC_I, SD
Destination
============
OS server ==> HP-UX
DB Name ==> CHDB
Existg DB users ==> 20
Oracle version ==> Oracle 10.2.0.4
copied 9i dump file in ==> /u03/export
Tablespaces created same as Oracle 9i i.e;
Tablespaces ==> SC, SC_I, SA, SA_I, PP, PP_I, AC, AC_I, SD
I had run full import on Destination server i.e; Oracle 10g & following is the command i used:
imp system/pwd@conn_string file=/u03/export/Jan09.dmp log=/u03/export/Jan09_imp.log full=y ignore=y statistics=none
Imported successfully with few errors like
Now my question is
===================
How to check everything is same/similar like Oracle 9i database ( source ), for example ( i used this technique ):
What are the things to check after full imp in 10g?
MSST_DB>SHOW USER
MSST_DB>SA
MSST_DB>select object_type, count(*) from user_objects group by objects;
OBJECT_TYPE COUNT(*)
DATABASE LINK 2
FUNCTION 23
INDEX 1795
LOB 6
PACKAGE 8
PACKAGE BODY 8
PROCEDURE 30
SEQUENCE 67
SYNONYM 60
TABLE 644
TRIGGER 3
VIEW 20
CHDB_DB>SHOW USER
CHDB_DB>SA
CHDB_DB>select object_type, count(*) from user_objects group by objects;
OBJECT_TYPE COUNT(*)
INDEX 1794
PROCEDURE 30
TABLE 644
TRIGGER 3
VIEW 20
FUNCTION 23
SYNONYM 60
PACKAGE BODY 8
SEQUENCE 67
PACKAGE 8
LOB 6

What are the things to check after full imp in 10g?Only public database links which are in use by your these users which you export/import (Only if you did user level export/import and not DB level).
If you log files does not show error during creation of any object, you don't need to double check any thing. Otherwise do on with the method which you mentioned by counting the objects.
Or you can create a database link from source to destination and use a query with MINUS to find out if there is any object missing
select object_name,object_type from user_objects
MINUS
select object_name,object_type from user_objects@destination_database;Salman

Checks after phase MAIN_SHDRUN/ACT_UPG were negative!

Dear Guys,
We are facing issue during the upgrade from 4.6c to ECC 6.0 EHP4 in pre-processing phase,
Checks after phase MAIN_SHDRUN/ACT_UPG were negative!
Last error code set: Detected 257 errors summarizes in 'ACTUPG.ELG' Check logfiles 'ACTUPG.ELG' and 'I:\usr\sap\DV1\upg\abap\log\SAPup.ECO'
ACTUPG.ELG:
******* LIST OF ERRORS AND RETURN CODES ******* ************************************************** ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ DDIC ACTIVATION ERRORS and RETURN CODE in SAPA-60016INECCDIMP.DV1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1 ETP111 exit code : "8" ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ DDIC ACTIVATION ERRORS and RETURN CODE in SAPA-60016INERECRUIT.DV1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1 ETP111 exit code : "8" ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ DDIC ACTIVATION ERRORS and RETURN CODE in SAPA-60016INFICA.DV1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1 ETP111 exit code : "8" ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ DDIC ACTIVATION ERRORS and RETURN CODE in SAPA-60016INFICAX.DV1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1 ETP111 exit code : "8" ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ DDIC ACTIVATION ERRORS and RETURN CODE in SAPA-60016INFINBASIS.DV1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1 ETP111 exit code : "8" ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ DDIC ACTIVATION ERRORS and RETURN CODE in SAPA-60016ININSURANC.DV1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1 ETP111 exit code : "8" ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ DDIC ACTIVATION ERRORS and RETURN CODE in SAPA-60016INISM.DV1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1 ETP111 exit code : "8" ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ DDIC ACTIVATION ERRORS and RETURN CODE in SAPA-60016INISOIL.DV1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1 ETP111 exit code : "8" ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ DDIC ACTIVATION ERRORS and RETURN CODE in SAPA-60016INISPSCA.DV1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1 ETP111 exit code : "8" ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ DDIC ACTIVATION ERRORS and RETURN CODE in SAPA-60016INISUT.DV1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1 ETP111 exit code : "8" ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ DDIC ACTIVATION ERRORS and RETURN CODE in SAPA-60017INECCDIMP.DV1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1 ETP111 exit code : "8" ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ DDIC ACTIVATION ERRORS and RETURN CODE in SAPA-60017INERECRUIT.DV1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
......... Please help us to resolve the problem.
Regards,
Bharathi

Hi Bharathi,
Can you please provide the contents of the activation log SAPA-60016INECCDIMP.DV1 which provides the information about the error.
Please check the activation log SAPAAA*.DV1 to know which tables are not activated.
Thank You.
Regards,
Deepika

Performance check after CPU

Similar Messages

Maybe you are looking for