SAP servers slow performance issues

[Former Member]

Hi,

I have SAP R/3 4.7 on HP-UX 11.23 & Oracle 9i. Users are facing slow performance issues while working on these SAP servers...what are the checks at the Oracle Database level, if there is any tuning can be done to improve the overall performance. Please Help.

Its 1TB of Database.

Hit Ratio is above 95 %, Swaps - 0%

CPU Idle time above 30-40%

Response Time - 6-10 ms

When I tried to run DB Check through DB16, I got some warnings & Errors...

IN_WRONG_TABLESPACE

MISSING_INDEX

MISSING_STATISTICS

TABLESPACE_FULL

ARCHIVE_TOO_OLD

BACKUP_TOO_OLD

DB_WRITER_PROCESSES

LOG_CHECKPOINT_INTERVAL

OPTIMIZER_FEATURES_ENABLE

01555

TSP_FULL

60

Checkpoint not complete

Thanx

Rupali S

[Former Member · Accepted solution]

Hi Rupali,

I think you need to address the DB16 warnings ASAP as it is directly related to performance of R/3 system.

Fix the below issues ASAP,

1. Missing indexes - Regenrate the indexes for the table which is mentioned in DB16 using SE11

2. Missing statistics - Schedule the analyze and optimizer jobs using DB13 once in every week.

3. Tablespace full - If any of the table space has crossed above 90%, immediately extend the tablespace....if it is above 80% , check for fragmentation and perform a reorg.

4. Archive too old - Check the saparch directory and perform a backup & delete the old archive files...

Address these above issues ASAP and I think the system performance will drastically increase....you need to keep an eye on DB16 warnings on daily basis....

regards,

Vinodh.

[Former Member]

Rupali,

First please make sure you follow the standard Oracle Parameter base on notes 124361.

Second, please check the archive file interval, Rule of Tumb, the interval between one archive and other archive file should not lower than 1 minutes.

You can check the archive log file created time. and query the v$LOG from ST04 - Detail Analyst menu - Display V$ Table - V$LOG

In here your should monitor the status column.

If you can't found any INACTIVE status, its mean your log file not big enough to handle the load.

To solve that problem, you can increase the log file size or add more group.

For your reference, you can read the notes below:

618868 FAQ Oracle Performance

Regards,

Fendi Suyanto