Showing Total Memory Used Per Oracle DB Instance

The following command gets the output of "ps aux"...

... takes only the lines regarding Oracle DB ("ora_") ...

... and sums up the RSS (resident set size, the non-swapped physical memory that a task has used (in kiloBytes)) ...

... grouped by DB SID

ps aux --sort -rss | awk '/USER +PID | ora_/{ split($11,a,"_"); db=a[3]; if(db!="" && db!="RSS"){ m[db]+=$6; } } END{for (db in m){ printf("DB:%9s uses %7.1f MB RAM\n",db,m[db]/1024); tot+=m[db]} printf("%12sTotal %7.1f GB RAM\n"," ",tot/(1024*1024))}'

Output Example

(DB "Dfug" is a temporary DB that has been automatically created by RMAN in order to make a TSPITR)

DB:     Dfug uses   884.9 MB RAM

DB: C0201Z01 uses  4196.6 MB RAM

            Total     5.0 GB RAM

Number Comparison Statements in KSH (Korn-Shell)

Before showing two variants of a number comparison statement, I want to show an alternative method counting the occurrence of a string in another one.

The following "grep -c" counts the occurrences of ":" in the echoed string:

echo "gu:gus:bla:bla" | grep -c ":"

Output

3

Another way to write the statement from above is:

grep -c ":" <<< "gu:gus:bla:bla"

Output

3

Here two examples of similar number comparison statements using the output of the statement from above:

OS> if (( $(grep -c ":" <<< "gugus:blabla") == 1 )); then echo "TRUE"; else echo "FALSE"; fi

TRUE

OS> if (( $(grep -c ":" <<< "gugus blabla") == 1 )); then echo "TRUE"; else echo "FALSE"; fi

FALSE

OS> if [[ $(grep -c ":" <<< "gugus:blabla") -eq 1 ]]; then echo "TRUE"; else echo "FALSE"; fi

TRUE

OS> if [[ $(grep -c ":" <<< "gugus blabla") -eq 1 ]]; then echo "TRUE"; else echo "FALSE"; fi

FALSE

I consider the first variant to be more elegant.

But if you are mixing character and numerical comparisons, you will need to use the 2nd variant with "[[ ... ]]"

For example:

X="Hello"; if [[ "${X}" == "Hello" && $(grep -c ":" <<< "gugus:blabla") -eq 1 ]]; then echo "TRUE"; else echo "FALSE"; fi

TRUE

X="Bella"; if [[ "${X}" == "Hello" && $(grep -c ":" <<< "gugus:blabla") -eq 1 ]]; then echo "TRUE"; else echo "FALSE"; fi

FALSE

The following syntax gives wrong results

OS> X="Xello"; if (( "${X}" == "Hello" && $(grep -c ":" <<< "gugus:blabla") == 1 )); then echo "TRUE"; else echo "FALSE"; fi

TRUE

OS> X="Bella"; if (( "${X}" == "Hello" && $(grep -c ":" <<< "gugus:blabla") == 1 )); then echo "TRUE"; else echo "FALSE"; fi

TRUE <== Wrong

Matematical Syntax	"In Words"
==	-eq
<	-lt
>	-gt
=<	-le
>=	-ge
!=	-ne