Different SAP mount points reaching to 99% full

[Former Member]

Hi Masters,

Presently i am working on different Solaris implemented SAP Clients.

o/s --> Solaris

DB --> Oracle 10g

after reaching the threashold value server is throwing an alert message

say suppose

/Oracle 99% full

/USR/SAP 98% full

so what are the area do we have to look in to resoleve this issue specially on Unix platform .

[Former Member]

Hi,

/oracle ===> Have you a separate filesystem for oraarch ? .......How big is your DB alertlog..

/usr/sap ===> look for files TRACE* core dev_* stat files

Mark

[Former Member]

Hi ,

It seems that standard housekeeping SAP jobs are not running in your system.Kindly schedule all SAP standard jobs so that unneccesary job logs get deleted & you may overcome from thsi issue.

Thanks..

Mohit

[Former Member]

Is not standard jobs joblogs etc under global which is /sapmnt... on UNIX ?

Mark

[Former Member]

Get into the filesystem /oracle or /usr/sap

Use commands like du -sg * , to identify which directory is taking the major diskspace

get into the directory , use commands like df -g . to identify where is the dir mounted to example (/usr/sap/trans , see whether its mounted to /usr/sap or is it a separate mount on its own.

basically identify dir which are mounted to /usr/sap and /oracle and are using the major disk space .

Major areas to look at /usr/sap/SID/DV*/work , look for core files and other logs there.

Do the same with /oracle , see whether /oracle/stage is mounted onto it and see what do u have in it , try to find something which you can remove from there..

See whether you still have old oracle binaries remaining even after upgrades .

These are some starting points .

Dont underestimate the fact that you might need to actually increase the size of the filesystem

All the best,

Gerard

[sunny_pahuja2]

Hi,

you can try to delete some logs. For more information, check SAP Note 16513 - File system is full - what do I do

Thanks

Sunny