DB: Oracle
I'm accepting filesystem name from user input and would like to use that value in sql query to generate dynamic script to restore datafile.
I can generate dynamic script if I know what filesystems name will be but trying to figure out on how to take user input value and put that in sql query?
Runtime: ./gen_query.ksh -oldDB db1 -newDB db2 -mt /u01, /u02, /u03, /u04
values for -oldDB -newDB -mt are entered by users.
-mt values will be separated by comma ","
-mt values can differ from user and can be only two mounts or can be three or four and can start with anything (e.g. /u01 or /u06 etc) so I can't really hardcode them.
I think what I need is to chop -mt values and store them into another variable and use those variable to generate query.
Is such thing possible?
Dynamic query:
#!/bin/ksh
LOG_FILE = test_gen_query.log
exec >> $LOG_FILE 2>&1
gen_query=
SELECT 'set newname for datafile '''
|| file_name
|| ''' to ''/u01 or /u02 or /u03 or /u04'
|| replace to_char(mod(rownum, 4)
|| replace(substr(file_name, 5, length (file_name)),'$2','$4')
|| ''';'
FROM (
SELECT file_name, bytes from dba_data_files
ORDER BY bytes desc
)
/
Sample output required:
set newname for datafile '/u40/oradata/db1/test1.dbf' to '/u01/oradata/db2/test1.dbf';
set newname for datafile '/u40/oradata/db1/test2.dbf' to '/u02/oradata/db2/test2.dbf';
set newname for datafile '/u41/oradata/db1/test3.dbf' to '/u03/oradata/db2/test3.dbf';
set newname for datafile '/u40/oradata/db1/test4.dbf' to '/u04/oradata/db2/test4.dbf';
set newname for datafile '/u40/oradata/db1/test5.dbf' to '/u01/oradata/db2/test5.dbf';
set newname for datafile '/u40/oradata/db1/test6.dbf' to '/u02/oradata/db2/test6.dbf';
...
...
When you want to use arguments with a minus-sign, see
learn about getopts. When you are lazy, you can use fixed positions:
./gen_query.ksh db1 db2 /u01 /u02 /u03 /u04
and get the vars with:
if [ $# -lt 3 ]; then
echo "Usage: $0 olddb newdb mount(s)"
exit 1
fi
# assign the first 2 vars
olddb=$1
newdb=$2
# Move "pointer" in the arglist
shift 2
# Split remaining args on a space
for mnt in $*; do
echo "my sql for dbold ${olddb} to dbnew ${newdb} for mount ${mnt}"
done
Related
I am trying to get the LAG between Primary & Standby database using the below shell script. The query works fine returning the values "DATABASE IS OUTOFSYNC" or "DATABASE IS INSYNC" for an instance that has 1 Node which returns a single value, but I get an error "[: 0 1: integer expression expected" for an instance that has two Nodes which returns two values for the LAG on the first Node and the Second Node.
So here is the code:
#!/bin/bash
get_status=$(sqlplus -s "/as sysdba" <<EOF
set pagesize 0 feedback off verify off heading off echo off;
SELECT prim.seq - tgt.seq seq_gap
FROM
(
SELECT thread#, MAX(sequence#) seq, MAX(completion_time) tm
FROM
v\$archived_log
GROUP BY
thread#
)
prim,
(
SELECT thread#, MAX(sequence#) seq, MAX(completion_time) tm
FROM
v\$archived_log
WHERE
dest_id IN
(
SELECT
dest_id
FROM
v\$archive_dest
WHERE
target = 'STANDBY'
)
AND
applied = 'YES'
GROUP BY
thread#
)
tgt
WHERE
prim.thread# = tgt.thread#;
exit;
EOF
)
if [ "$get_status" -ge 5 ]; then
echo "DATABASE IS OUTOFSYNC"
else
echo "DATABASE IS INSYNC"
fi
Is there a better way to write this script?
After adding typeset -p get_status after the query and before the if I get the below results:
declare -- get_status=" 1
0"
./dgtest2.sh: line 41: [: 1
0: integer expression expected
DATABASE IS INSYNC
The query is returning more than one value/string (for 2 nodes or threads) as shown in picture/screenshot and it seems like my script is only coded to address a single value/string generated by the query.
enter image description here
Is there away to modify the script to address multiple values/strings generated by the query
The logic should be if all values returned are -ge 5 it should report "DATABASE IS OUTOFSYNC" else "DATABASE IS INSYNC" for all values returned are -lt 5.
The logic for one value -lt 5 and one value -ge 5 would not suffice as the values constantly change on the database.
Any values from 0 - 4 that the database returns whether from both Nodes should report as "DATABASE IS INSYNC" and any value from 5 upwards that the database returns whether from both Nodes should report as "DATABASE IS OUTOFSYNC".
One idea would be to capture the status values (returned by the sqlplus script) into an array and then loop through the array testing said status values.
Instead of:
variable=$(sqlplus ...)
We want:
variable=( $(sqlplus ...) )
For OP's current scripting, with a name change for the variable, we will replace this:
get_status=$(sqlplus -s "/as sysdba" <<EOF
set pagesize 0 feedback off verify off heading off echo off;
SELECT prim.seq - tgt.seq seq_gap
...
exit;
EOF
)
With this:
status_array=( $(sqlplus -s "/as sysdba" <<EOF
set pagesize 0 feedback off verify off heading off echo off;
SELECT prim.seq - tgt.seq seq_gap
...
exit;
EOF
) )
One idea for the follow-on logic testing:
default database status is INSYNC
if any status values are -ge 5 then set database status to OUTOFSYNC
The code for this looks like:
db_status='INSYNC'
for status in "${status_array[#]}"
do
[[ "${status}" -ge 5 ]] && db_status='OUTOFSYNC' && break
done
echo "DATABASE IS ${db_status}"
I'm not setup to run the sqlplus script but I should be able to simulate the results with the following array assignments:
status_array=(1)
status_array=(7)
status_array=(0 1)
status_array=(5 7)
status_array=(5 3)
Running our code for each of these array assignments gives us:
##################### status_array=(1)
DATABASE is INSYNC
##################### status_array=(7)
DATABASE is OUTOFSYNC
##################### status_array=(0 1)
DATABASE is INSYNC
##################### status_array=(5 7)
DATABASE is OUTOFSYNC
##################### status_array=(5 3)
DATABASE is OUTOFSYNC
I have a function like so -
CREATE
OR REPLACE FUNCTION ind (bucket text) RETURNS table (
middle character varying (100),
last character varying (100)
) AS $body$ BEGIN return query
select
fname as first,
lname as last
from all_records
; END;
$body$ LANGUAGE PLPGSQL;
How do I output the results of select ind ('Mob') into a tsv file?
I want the output to look like this -
first last
MARY KATHERINE
You can use the COPY command
example:
COPY (select * from ind('Mob')) TO '/tmp/ind.tsv' CSV HEADER DELIMITER E'\t';
the file '/tmp/ind.tsv' will contain you data
Postgres doesn't allow copy with header for tsv for some reason.
If you're using a linux based system you can do it with a script like this:
#create file with tab delimited column list (use \t between each column name)
echo -e "user_id\temail" > user_output.tsv
#now you can append the results of your query to that file by copying to STDOUT
psql -h your_host_name -d your_database_name -c "\copy (SELECT user_id, email FROM my_user_table) to STDOUT;" >> user_output.tsv
Alternatively, if your script is long and you don't want to pass it in with -c command you can use the same approach from a .sql file, use "--quiet" to avoid notices being passed into your file
psql --quiet -h your_host_name -d your_database_name -f your_sql_file.sql >> user_output.tsv
My code is:
#!/bin/sh
cat tmp_ts.log | awk ' {print $8}'
lookup=$8
sqlplus -s "sys/Orcl1234 as sysdba" << EOF
SELECT tablespace_name FROM dba_tablespaces WHERE tablespace_name='$lookup';
exit;
EOF
and my output is:
IAM_OIM
no rows selected
In this variable lookup I have passed to select statement but it's not working.
My end result should be with select statement. See below the output of select query:
See below:
My end result should be this but that variable is not working in select statement.
#!/bin/sh
lookup="$(awk '/tablespace/{print $8;exit}' tmp_ts.log)"
echo "Querying database with lookup = $lookup"
sqlplus -s "sys/Orcl1234 as sysdba" <<EOF
SELECT tablespace_name FROM dba_tablespaces WHERE tablespace_name='$lookup';
exit;
EOF
You have to use awk's output to set lookup. The shell knows nothing about the $8 which was set in awk. Also, I have ensured that awk exits after the first matching line, so that there is no risk of returning multiple values, or simply empty lines as it did in your version.
You can fill lookup with a command like awk, sed or cut.
lookup=$(cut -d" " -f8 tmp_ts.log)
You should add some checks, like #Dario did (with an exit after the first match and only converting lines with tablespace, but what to do when no lines match?).
When you don't add the checks you can skip setting the $lookup:
sqlplus -s "sys/Orcl1234 as sysdba" << EOF
SELECT tablespace_name FROM dba_tablespaces
WHERE tablespace_name='$(sed 's/.*tablespace- //' tmp_ts.log)';
exit;
EOF
I want to pass shell variables to sql statement. Both shell script and SQL statement are present in the same script file.
I want the values of the variables retMonth, retLastDay and retPrvYear in the SQL statement.
Below is the code.
If I execute this, it prints - " partition_date between '01--' and '--' \ 0 0] 1 1] 12-DEC-14 1"
How can I have values of retMonth, retLastDay and retPrvYear in SQL statement?
echo $retMonth //This prints 07
echo $retLastDay //This prints 31
echo $retPrvYear //This prints 2015
count=$(sqlplus -s ${DBA_ORACLE_USER}/${DBA_ORACLE_PWORD}#${ORACLE_SID} <<END
#connect ${DBA_ORACLE_USER}/${DBA_ORACLE_PWORD}#${ORACLE_SID}
set serveroutput on
set linesize 1000
set heading off
set feedback off
define lastMonth=$retMonth
define lastYear=$retPrvYear
define lastDay=$retLastDay
SELECT count(1)
FROM MYTABLE
WHERE partition_date between '01-$lastMonth-$lastYear' and '$lastDay-$lastMonth-$lastYear'
);
END
)
Try using quoted shell variables directly without using define directives:
count=$(sqlplus -s "${DBA_ORACLE_USER}/${DBA_ORACLE_PWORD}#${ORACLE_SID}" <<END
set serveroutput on
set linesize 1000
set heading off
set feedback off
SELECT count(1)
FROM MYTABLE
WHERE partition_date between
"01-$retMonth-$retPrvYear" and "$retLastDay-$retMonth-$retPrvYear";
END
)
I have a select statement inside a bash script that returns the latest date in the DB. I run this query 4 times so I want to define it just once and assing the text to a variable.
#!/bin/bash
linux commands;
database_date=$(sqlplus -s/nolog $USER/$USER#BRMDPP <<END
set pagesize 0 feedback off verify off heading off echo off;
SELECT ...
exit;
END
)
commands that change the database date;
last_date=$(sqlplus -s/nolog $USER/$USER#BRMDPP <<END
set pagesize 0 feedback off verify off heading off echo off;
SELECT ...
exit;
END
)
commands that change the database date;
How can I store this big string $(sqlplus ... into one variable and use it again?
Thank you
One way would be to make use of a function:
foo() {
sqlplus -s/nolog $USER/$USER#BRMDPP <<END
set pagesize 0 feedback off verify off heading off echo off;
SELECT ...
exit;
END
}
and later invoke it by saying:
value=$(foo)
In order to get the value returned by the function, say echo "$value" (note that quoting variables is important).