TOP SQL reporting
---- mon_topsql.sql
https://github.com/vishaldesai/Oracle_Scripts/blob/master/Exadata/mon_topsql.sql
-- Author: Tanel Poder ( http://blog.tanelpoder.com | tanel@tanelpoder.com )
SET LINES 999 PAGES 5000 TRIMSPOOL ON TRIMOUT ON TAB OFF
COL pct FOR A10 JUST RIGHT
COL cpu_pct FOR 999.9
COL io_pct FOR 999.9
col OWNER for a20
col OBJECT_NAME for a20
col PROCEDURE_NAME for a20
BREAK ON day SKIP 1
DEF days=1
PROMPT Displaying daily top SQL for last &days days...
WITH ash AS (
SELECT
day
, owner
, object_name
, procedure_name
, sql_id
, sql_plan_hash_value
, total_seconds
, io_seconds
, cpu_seconds
, LPAD(TRIM(TO_CHAR(RATIO_TO_REPORT(total_seconds) OVER (PARTITION BY day) * 100, '999.9'))||'%', 10) pct
, RATIO_TO_REPORT(total_seconds) OVER (PARTITION BY day) * 100 pct_num
FROM (
SELECT
TO_CHAR(sample_time, 'YYYY-MM-DD') day
, sql_id
, sql_plan_hash_value
, p.owner
, p.object_name
, p.procedure_name
, SUM(10) total_seconds
, SUM(CASE WHEN wait_class = 'User I/O' THEN 10 ELSE 0 END) io_seconds
, SUM(CASE WHEN wait_class IS NULL THEN 10 ELSE 0 END) cpu_seconds
FROM
dba_hist_active_sess_history a
, dba_procedures p
WHERE
a.plsql_entry_object_id = p.object_id (+)
AND a.plsql_entry_subprogram_id = p.subprogram_id (+)
AND sample_time > SYSDATE - &days
AND session_type != 'BACKGROUND' -- ignore for now
GROUP BY
sql_id
, sql_plan_hash_value
, p.owner
, p.object_name
, p.procedure_name
, TO_CHAR(sample_time, 'YYYY-MM-DD')
)
)
, sqlstat AS (
SELECT /*+ MATERIALIZE */
TO_CHAR(begin_interval_time, 'YYYY-MM-DD') day
, sql_id
, plan_hash_value
, SUM(executions_delta) executions
, SUM(rows_processed_delta) rows_processed
, SUM(disk_reads_delta) blocks_read
, SUM(disk_reads_delta)*8/1024 mb_read
, SUM(buffer_gets_delta) buffer_gets
, SUM(iowait_delta)/1000000 awr_iowait_seconds
, SUM(cpu_time_delta)/1000000 awr_cpu_seconds
, SUM(elapsed_time_delta)/1000000 awr_elapsed_seconds
FROM
dba_hist_snapshot
NATURAL JOIN
dba_hist_sqlstat
WHERE
begin_interval_time > SYSDATE - &days
GROUP BY
TO_CHAR(begin_interval_time, 'YYYY-MM-DD')
, sql_id
, plan_hash_value
)
SELECT /*+ MONITOR */
day
, pct
, owner
, object_name
, procedure_name
, sql_id
, sql_plan_hash_value plan_hash
, ROUND(total_seconds / 3600,1) total_hours
, total_seconds
, executions
, ROUND(total_seconds / NULLIF(executions,0),2) seconds_per_exec
, io_pct
, cpu_pct
, mb_read
, ROUND(mb_read / NULLIF(executions,0),2) mb_per_exec
, buffer_gets
, ROUND(buffer_gets / NULLIF(executions,0),2) bufget_per_exec
, CASE WHEN sql_id IS NOT NULL THEN
'SELECT * FROM TABLE(DBMS_XPLAN.DISPLAY_AWR('''||sql_id||''','||CASE WHEN sql_plan_hash_value = 0 THEN 'NULL' ELSE TO_CHAR(sql_plan_hash_value) END||', format=>''ADVANCED''));'
END extract_plan_from_awr
FROM (
SELECT
day
, pct
, owner
, object_name
, procedure_name
, sql_id
, sql_plan_hash_value
, total_seconds
, io_seconds/total_seconds*100 io_pct
, cpu_seconds/total_seconds*100 cpu_pct
, (SELECT executions FROM sqlstat s WHERE ash.sql_id = s.sql_id AND ash.sql_plan_hash_value = s.plan_hash_value AND ash.day = s.day) executions
, (SELECT mb_read FROM sqlstat s WHERE ash.sql_id = s.sql_id AND ash.sql_plan_hash_value = s.plan_hash_value AND ash.day = s.day) mb_read
, (SELECT buffer_gets FROM sqlstat s WHERE ash.sql_id = s.sql_id AND ash.sql_plan_hash_value = s.plan_hash_value AND ash.day = s.day) buffer_gets
FROM
ash
WHERE
ash.pct_num >= 1
)
ORDER BY
day DESC
, total_seconds DESC
/
DAY PCT OWNER OBJECT_NAME PROCEDURE_NAME SQL_ID PLAN_HASH TOTAL_HOURS TOTAL_SECONDS EXECUTIONS SECONDS_PER_EXEC IO_PCT CPU_PCT MB_READ MB_PER_EXEC BUFFER_GETS BUFGET_PER_EXEC EXTRACT_PLAN_FROM_AWR
---------- ---------- -------------------- -------------------- -------------------- ------------- ---------- ----------- ------------- ---------- ---------------- ------ ------- ---------- ----------- ----------- --------------- --------------------------------------------------------------------------------------------------------------------------
2023-08-12 78.6% ctyc1af1abg5u 200298931 91 327510 .0 100.0 SELECT * FROM TABLE(DBMS_XPLAN.DISPLAY_AWR('ctyc1af1abg5u',200298931, format=>'ADVANCED'));
4.4% 0 5.1 18340 22.3 72.7
4.2% 3xjw1ncw5vh27 2487977020 4.9 17500 88.3 11.7 SELECT * FROM TABLE(DBMS_XPLAN.DISPLAY_AWR('3xjw1ncw5vh27',2487977020, format=>'ADVANCED'));
4.2% c7ngymcfjtb1c 1658157256 4.8 17450 43.8 56.2 SELECT * FROM TABLE(DBMS_XPLAN.DISPLAY_AWR('c7ngymcfjtb1c',1658157256, format=>'ADVANCED'));
2023-08-11 55.1% ctyc1af1abg5u 200298931 55.9 201130 .0 100.0 SELECT * FROM TABLE(DBMS_XPLAN.DISPLAY_AWR('ctyc1af1abg5u',200298931, format=>'ADVANCED'));
8.7% 1zuyhdtsmmx5s 3008400300 8.8 31680
define sql_id='ctyc1af1abg5u'
SET TERMOUT OFF pagesize 5000 tab off verify off linesize 999 trimspool on trimout on null ""
SET TERMOUT ON
COL exec_per_sec FOR 99999990
COL ela_ms_per_sec FOR 99999990
COL rows_per_sec FOR 99999990
COL lios_per_sec FOR 99999990
COL blkrd_per_sec FOR 99999990
COL cpu_ms_per_sec FOR 99999990
COL iow_ms_per_sec FOR 99999990
COL clw_ms_per_sec FOR 99999990
COL apw_ms_per_sec FOR 99999990
COL ccw_ms_per_sec FOR 99999990
SELECT
CAST(begin_interval_time AS DATE) begin_interval_time
,st.con_id
, sql_id
, plan_hash_value
, ROUND(SUM(executions_delta ) / ((CAST(end_interval_time AS DATE) - CAST(begin_interval_time AS DATE)) * 86400), 1) exec_per_sec
, ROUND(SUM(elapsed_time_delta ) / 1000 / ((CAST(end_interval_time AS DATE) - CAST(begin_interval_time AS DATE)) * 86400), 1) ela_ms_per_sec
, ROUND(SUM(rows_processed_delta) / ((CAST(end_interval_time AS DATE) - CAST(begin_interval_time AS DATE)) * 86400), 1) rows_per_sec
, ROUND(SUM(buffer_gets_delta ) / ((CAST(end_interval_time AS DATE) - CAST(begin_interval_time AS DATE)) * 86400), 1) lios_per_sec
, ROUND(SUM(disk_reads_delta ) / ((CAST(end_interval_time AS DATE) - CAST(begin_interval_time AS DATE)) * 86400), 1) blkrd_per_sec
, ROUND(SUM(cpu_time_delta ) / 1000 / ((CAST(end_interval_time AS DATE) - CAST(begin_interval_time AS DATE)) * 86400), 1) cpu_ms_per_sec
, ROUND(SUM(iowait_delta ) / 1000 / ((CAST(end_interval_time AS DATE) - CAST(begin_interval_time AS DATE)) * 86400), 1) iow_ms_per_sec
, ROUND(SUM(clwait_delta ) / 1000 / ((CAST(end_interval_time AS DATE) - CAST(begin_interval_time AS DATE)) * 86400), 1) clw_ms_per_sec
, ROUND(SUM(apwait_delta ) / 1000 / ((CAST(end_interval_time AS DATE) - CAST(begin_interval_time AS DATE)) * 86400), 1) apw_ms_per_sec
, ROUND(SUM(ccwait_delta ) / 1000 / ((CAST(end_interval_time AS DATE) - CAST(begin_interval_time AS DATE)) * 86400), 1) ccw_ms_per_sec
FROM
dba_hist_snapshot sn
, dba_hist_sqlstat st
WHERE
sn.snap_id = st.snap_id
AND sn.dbid = st.dbid
AND sn.instance_number = st.instance_number
AND sql_id = '&sql_id'
--AND plan_hash_value LIKE '2'
--AND begin_interval_time >= 3
--AND end_interval_time <= 4
-- AND begin_interval_time > sysdate -1
GROUP BY
CAST(begin_interval_time AS DATE)
,st.con_id
, CAST(end_interval_time AS DATE)
, sql_id
, plan_hash_value
ORDER BY
begin_interval_time
, sql_id
, plan_hash_value
/
SET LINESIZE 200 PAGESIZE 200 TRIMSPOOL ON
COLUMN username FORMAT a20
COLUMN total_WaitTime FORMAT 999,999.99 HEADING "WAIT_SEC"
COLUMN comment FORMAT a30
COLUMN inst_id FORMAT 9999
COLUMN sid FORMAT 999999
col kill for a17
col EVENT for a20
SELECT t.*,
CASE
WHEN ROWNUM <= 5 THEN 'Highest wait: ' || t.username || ' with ' || t.total_WaitTime || ' seconds'
ELSE 'Waits by user last 15 minutes'
END AS "comment"
FROM (
SELECT
''''||s.sid ||','|| s.serial#||',@'||s.inst_id ||'''' kill,
--s.inst_id,
-- s.sid,
s.con_id,
s.username,
SUM(ash.wait_time + ash.time_waited) / 1000000 AS total_WaitTime,
s.sql_id,
s.event
FROM gv$active_session_history ash
JOIN gv$session s ON ash.session_id = s.sid AND ash.inst_id = s.inst_id
WHERE 1=1
-- ash.sample_time > (SYSDATE - 1/96) -- Last 15 minutes
and ash.sample_time > SYSDATE - INTERVAL '15' MINUTE
AND s.username IS NOT NULL
GROUP BY ''''||s.sid ||','|| s.serial#||',@'||s.inst_id ||'''',s.con_id, s.username,s.sql_id,s.event
ORDER BY 4 DESC
) t
WHERE ROWNUM < 20;
from https://www.pythian.com/blog/recent-spike-report-from-vactive_session_history-ash
AVG — the average “load” (active sessions) over a 5-minute interval. This should help you spot a problem when you scroll through the results.
TOT — total “load” (active sessions) for that sample time. RAC users: each RAC node will have its own sample time, within 1 second of each other, but not exactly spot-on. So, even if you have sessions waiting on the same event, they will not be grouped together. I kind of like it this way, for now.
SAMPLE_TIME — self-explanatory
CNT — the number of active sessions waiting on the same event and query
EVENT — the event been waited on
SQL_TEXT — self-explanatory, except when empty which means either not found in shared pool or not available in ASH
SQL_ID — if you need to find the SQL
CHD — the child number being executed
COLUMN event FORMAT a25 TRUNCATED
COLUMN "SUBSTR(SQ.SQL_TEXT,1)" FORMAT a40 HEADING "SQL_TEXT" TRUNCATED
col SAMPLE_TIME for a25
-- Metric Columns
COLUMN "sec p" FORMAT a15
COLUMN "disk p" FORMAT 999,999
COLUMN "gets p" FORMAT 9,999,999
COLUMN "rows p" FORMAT 999,999
COLUMN "cpu p" FORMAT 99.999
COLUMN exec FORMAT 999,999,999
COLUMN open FORMAT 999
COLUMN e FORMAT 999
col SQL_TEXT for a50
set linesize 300 pagesize 300
SELECT
ROUND(AVG(MAX(cnt_tot)) OVER (ORDER BY sample_time RANGE BETWEEN INTERVAL '5' MINUTE PRECEDING AND CURRENT ROW)) AS avg,
MAX(cnt_tot) AS tot,
sample_time,
MAX(cnt) AS cnt,
event,
SUBSTR(sq.sql_text, 1, 100) as sql_text, -- Truncated for display
ash.sql_id,
ash.sql_child_number AS chd,
TO_CHAR(ROUND(SUM(elapsed_time) / NULLIF(SUM(executions), 0) / 1000000, 6), '9,990.999999') AS "sec p",
ROUND(SUM(disk_reads) / NULLIF(SUM(executions), 0), 0) AS "disk p",
ROUND(SUM(buffer_gets) / NULLIF(SUM(executions), 0), 0) AS "gets p",
ROUND(SUM(rows_processed) / NULLIF(SUM(executions), 0), 0) AS "rows p",
ROUND(SUM(cpu_time) / 1000000 / NULLIF(SUM(executions), 0), 3) AS "cpu p",
SUM(executions) AS exec,
SUM(users_opening) AS open,
SUM(users_executing) AS e
FROM (
SELECT SUM(COUNT(*)) OVER (PARTITION BY sample_time) AS cnt_tot,
COUNT(*) AS cnt,
sample_time,
event,
sql_id,
sql_child_number
FROM gv$active_session_history
WHERE sample_time > SYSDATE - INTERVAL '1' HOUR
GROUP BY event, sql_id, sql_child_number, sample_time
HAVING COUNT(*) >= 10
) ash
LEFT JOIN gv$sql sq ON ash.sql_id = sq.sql_id
AND ash.sql_child_number = sq.child_number
GROUP BY event, sql_text, ash.sql_id, ash.sql_child_number, ash.sample_time
ORDER BY sample_time DESC;
connnect to pdbs
col PLAN_TABLE_OUTPUT for a150
SELECT * FROM TABLE(DBMS_XPLAN.DISPLAY_AWR('ctyc1af1abg5u',200298931, format=>'ADVANCED'));
Saturday, 12 August 2023
TOP SQL reporting
TOP SQL reporting
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment