由研究oracle rac lms程式引發10708 event及oradebug dump buffer
結論
1,對於報錯的分析，可見猜測能力非常重要，聯想能力極為重要這個非常重要，如果理解報錯，直接決定你分析思路，最終要花多久時間解決問題
2,oradebug用法還是理解不深
3,oradebug dump buffer 後面的數字，不是字面的級別，而是緩衝區中資料塊的地址即rdba
4,rdba可以基於測試表dbms_rowid,然後dump block,最後查詢rdba,並把16進位制的RDBA轉化為10進位制即可
5,oradebug dump buffer 10進位制的block_rdba
還有一個伴隨的報錯，要用一個事件
alter session set events 'immediate trace name set_tsn_p1 level 6';（也要根據報錯進行反推）
6,10708 event跟蹤RAC BUFFER CACHE，歷經各種測試，仍無果，
基於LMD,LMS以及使用者程式跟蹤，全無有價值的資訊
還有其它的引數沒有配置嗎，如果是，到底是什麼引數呢
或者是我採用ORACEBUG的命令或ALTER SYSTEM SET EVENTS的命令不對，到底怎麼回事呢
7，引申下，oradebug有些命令，不能僅看其命令的表面意思，一定要擴充套件思維，這個非常重要
也就是說有時會誤導大家，死活想不通，ORACLE就是這樣
8, oradebug dump events根據後面數字的不同，分為跟蹤當前會話，某個程式，資料庫系統中，這些不對級別配置的事件，並記錄到TRC檔案中
oradebug dump events後面的數字，有效值為：
1 當前會話
2 某個程式（區別於上面，比如你想跟蹤某個程式）
4 整個資料庫系統
9, 還是哪種老話，一定要好好理解或思考報錯的資訊，我對於oradebug dump events後面的數字理解，就是源於報錯資訊
10,用oradebug dump events 是有資料了，但只有條目，沒有真正的資料
這裡就需要猜測及聯想能力，說明我這塊對於ORACLE的理解以及應用仍然不足
這塊能力要提升上去，將是一個極大的提升分析能力的過程
11,繼續上述第6的分析，嘗試了oradebug,alter session，最後透過alter system set events即可讓10708跟蹤RAC BUFFER CACHE
整個過程還是比較艱辛的，加油
具體指令碼如下：
SQL> alter system set event='10708 trace name context forever,level 10' scope=spfile;
System altered.
SQL> create pfile='/home/oracle/10gpfile.txt' from spfile;
File created.
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
SQL> startup
ORACLE instance started.
Total System Global Area 599785472 bytes
Fixed Size 2022632 bytes
Variable Size 335545112 bytes
Database Buffers 260046848 bytes
Redo Buffers 2170880 bytes
Database mounted.
Database opened.
[oracle@jingfa1 udump]$ ll -lrht
-rw-r----- 1 oracle oinstall 65 Nov 12 19:10 jingfa1_ora_19482_hbstest.trc
-rw-r----- 1 oracle oinstall 65 Nov 12 19:12 jingfa1_ora_26927_hbstest.trc
-rw-r----- 1 oracle oinstall 1.5K Nov 12 19:37 jingfa1_ora_31562.trc
-rw-r----- 1 oracle oinstall 969 Nov 12 19:38 jingfa1_ora_1779.trc
-rw-rw---- 1 oracle oinstall 593 Nov 12 19:38 jingfa1_ora_2035.trc
-rw-r----- 1 oracle oinstall 759 Nov 12 19:38 jingfa1_ora_1981.trc
-rw-rw---- 1 oracle oinstall 593 Nov 12 19:38 jingfa1_ora_2234.trc
-rw-rw---- 1 oracle oinstall 593 Nov 12 19:38 jingfa1_ora_2233.trc
-rw-r----- 1 oracle oinstall 37K Nov 12 19:38 jingfa1_ora_2261.trc
-rw-rw---- 1 oracle oinstall 3.2K Nov 12 19:39 jingfa1_ora_2224.trc
[oracle@jingfa1 udump]$ date
kclmap: mapping region: 0x60400000 length: 595591168
*** 2015-11-12 19:39:39.982
*** SERVICE NAME:(SYS$USERS) 2015-11-12 19:39:39.982
*** SESSION ID:(146.1) 2015-11-12 19:39:39.981
kclscrs: req=0 block=1/400760
kclscrs: req=0 typ=immediate mode=S
kclscrs: req=0 block=1/400761
kclscrs: req=0 typ=immediate mode=S
kclscrs: req=0 block=1/400763
kclscrs: req=0 typ=immediate mode=S
中間略
kclscrs: req=0 block=1/4022ea
kclwcrs: req=0 typ=cur wtyp=2hop tm=749
kclscrs: req=0 block=3/c01633
kclwcrs: req=0 typ=stale
kclscrs: req=0 block=3/c016fc
kclwcrs: req=0 typ=cur wtyp=2hop tm=870
[oracle@jingfa1 udump]$
測試
SQL> select spid from v$process where addr=(select paddr from v$session where sid=130);
SPID
------------
29546
SQL> oradebug setospid 29546
Oracle pid: 22, Unix process pid: 29546, image: oracle@jingfa1 (TNS V1-V3)
SQL>
SQL> oradebug event 10708 trace name context forever,level 12;
Statement processed
SQL> create table t_rac_cross_buffer(a int);
Table created.
SQL> insert into t_rac_cross_buffer select * from t_rac_buffer;
10000 rows created.
SQL> commit;
Commit complete.
SQL> oradebug event 10708 trace name context off;
Statement processed.
SQL> oradebug tracefile_name
/u01/app/oracle/admin/jingfa/udump/jingfa1_ora_29546.trc
檢視TRACE FILE沒發現有任何資訊
Received ORADEBUG command 'event 10708 trace name context forever,level 12' from process Unix process pid: 7178, image:
*** 2015-11-12 16:02:49.680
Received ORADEBUG command 'event 10708 trace name context off' from process Unix process pid: 7178, image:
Received ORADEBUG command 'tracefile_name' from process Unix process pid: 7178, image
換個思路
仍然沒有資訊
採用oradebug dump buffer居然報錯
SQL> oradebug setmypid
Statement processed.
SQL> oradebug dump buffer 12
Statement processed.
SQL> oradebug flush
Statement processed.
SQL> oradebug tracefile_name
/u01/app/oracle/admin/jingfa/udump/jingfa1_ora_30018.trc
WARNING: set_tsn_p1 event has not been called.
Using stale value for tablespace number - 0
Dump of buffer cache at level 10 for tsn=0, rdba=12
SQL> alter session set events 'immediate trace name set_tsn_p1 level 12';
Session altered.
可見oradebug dump buffer跟蹤的數字對應表空間的編號，且這個資料字對應資料塊的地址RDBA
SQL> oradebug setmypid
Statement processed.
SQL> oradebug dump buffer 12
Statement processed.
SQL> oradebug flush
Statement processed.
SQL> oradebug tracefile_name
/u01/app/oracle/admin/jingfa/udump/jingfa1_ora_9512.trc
Dump of buffer cache at level 10 for tsn=11, rdba=12
即使這個數字是RDBA，我們換個測試表的RDBA進行測試
SQL> select dbms_rowid.rowid_relative_fno(rowid),dbms_rowid.rowid_block_number(rowid),dbms_rowid.rowid_row_number(rowid),a from t_rac_cross_buffer where a=661;
DBMS_ROWID.ROWID_RELATIVE_FNO(ROWID) DBMS_ROWID.ROWID_BLOCK_NUMBER(ROWID) DBMS_ROWID.ROWID_ROW_NUMBER(ROWID) A
------------------------------------ ------------------------------------ ---------------------------------- ----------
5 22724 0 661
5 22744 100 661
5 22760 200 661
5 22776 300 661
5 22792 400 661
5 26904 500 661
5 26920 600 661
7 rows selected.
SQL> alter session set tracefile_identifier='block_strtest';
Session altered.
SQL> alter system dump datafile 5 block 22724;
System altered.
[oracle@jingfa1 udump]$ more jingfa1_ora_29546_block_strtest.trc|grep rdba
buffer tsn: 5 rdba: 0x014058c4 (5/22724)
SQL> select dbms_utility.data_block_address_file(to_number('014058c4','xxxxxxxx')) file_no,dbms_utility.data_block_address_block(to_number('014058c4','xxxxxxxx')) block_no from dual;
FILE_NO BLOCK_NO
---------- ----------
5 22724
SQL> select ts#,name from v$tablespace;
TS# NAME
---------- ------------------------------
0 SYSTEM
1 UNDOTBS1
2 SYSAUX
3 TEMP
4 UNDOTBS2
5 USERS
SQL> alter session set events 'immediate trace name set_tsn_p1 level 4';--測試表空間編號減去1，由上述報錯可知
Session altered.
SQL> oradebug setmypid
Statement processed.
SQL> oradebug dump buffer 014058c4 --不能直接DUMP 測試表所對應的16進位制RDBA
ORA-01858: a non-numeric character was found where a numeric was expected
SQL> select to_number('014058c4','xxxxxxxx') from dual;--16進位制RDBA轉化為10進位制
TO_NUMBER('014058C4','XXXXXXXX')
--------------------------------
20994244
SQL> oradebug dump buffer 20994244
Statement processed.
SQL> oradebug tracefile_name
/u01/app/oracle/admin/jingfa/udump/jingfa1_ora_8888.trc
SQL>
有報錯
Dump of buffer cache at level 10 for tsn=3, rdba=20994244
SQL> alter session set events 'immediate trace name set_tsn_p1 level 6';
Session altered.
SQL> oradebug setmypid
Statement processed.
SQL> oradebug dump buffer 20994244
Statement processed.
SQL> oradebug tracefile_name
/u01/app/oracle/admin/jingfa/udump/jingfa1_ora_28767.trc
SQL>
*** 2015-11-12 17:05:12.795
*** SERVICE NAME:(SYS$USERS) 2015-11-12 17:05:12.794
*** SESSION ID:(141.769) 2015-11-12 17:05:12.794
Dump of buffer cache at level 10 for tsn=5, rdba=20994244
BH (0x6e3d95f8) file#: 5 rdba: 0x014058c4 (5/22724) class: 1 ba: 0x6e2fc000
set: 3 blksize: 8192 bsi: 0 set-flg: 2 pwbcnt: 290
dbwrid: 0 obj: 10447 objn: 10447 tsn: 5 afn: 5
hash: [82b7ae88,82b7ae88] lru: [6e7d1e68,6e7cf718]
lru-flags: on_auxiliary_list
ckptq: [NULL] fileq: [NULL] objq: [NULL]
st: FREE md: NULL tch: 0 le: (nil)
flags:
GLOBAL CACHE ELEMENT DUMP (address: 0x627e3a80):
id1: 0x58c4 id2: 0x50000 obj: 10447 block: (5/22724)
lock: NC rls: 0x0000 acq: 0x0000 latch: 0
flags: 0x43 fair: 0 recovery: 0 fpin: 'ktspbwh2: ktspfmdb'
bscn: 0x0.0 bctx: (nil) write: 0 scan: 0x0 xflg: 0 xid: 0x0.0.0
lcp: (nil) lnk: [NULL] lch: [0x627ea788,0x627e74c8]
seq: 15 hist: 197 48 1 66 144:0 192 352 197 48 1 113 424 143:0 325 352
GCS SHADOW 0x627e3af0,3 sq[0x627e3af0,0x627e3af0] resp[(nil),0x58c4.50000] pkey 10447
grant 0 cvt 0 mdrole 0x0 st 0x40 GRANTQ rl LOCAL
master 0 owner 0 sid 0 remote[(nil),0] hist 0x8230118c
history 0xc.0x6.0x1.0xc.0x2.0x2. cflag 0x0 sender 0 flags 0x0 replay# 0
disk: 0x0000.00000000 write request: 0x0000.00000000
pi scn: 0x0000.00000000
msgseq 0x0 updseq 0x0 reqids[2,0,0] infop 0x0
pkey 10447
hv 49 [stat 0x0, 0->0, wm 32767, RMno 0, reminc 0, dom 0]
kjga st 0x4, step 0.0.0, cinc 4, rmno 4, flags 0x0
lb 0, hb 0, myb 280, drmb 280, apifrz 0
再回過頭看下10708 EVENT
SQL> select spid,pid,program from v$process where lower(program) like '%lms%';
SPID PID PROGRAM
------------ ---------- ------------------------------------------------
5183 7 oracle@jingfa1 (LMS0)
SQL> oradebug setospid 5183
Oracle pid: 7, Unix process pid: 5183, image: oracle@jingfa1 (LMS0)
SQL> oradebug event 10708 trace name context forever,level 12;
Statement processed.
SQL> show user
USER is "TBS_ZXY"
SQL> create table t_boc(a int);
Table created.
SQL> insert into t_boc select level from dual connect by level<=10000;
10000 rows created.
SQL> commit;
Commit complete.
SQL> oradebug event 10708 trace name context off;
Statement processed.
SQL> oradebug tracefile_name
/u01/app/oracle/admin/jingfa/bdump/jingfa1_lms0_5183.trc
跟蹤了LMS，沒有看到有價值的資訊
*** SERVICE NAME:() 2015-11-12 15:46:29.138
*** SESSION ID:(165.1) 2015-11-12 15:46:29.138
0 GCS shadows cancelled, 0 closed
0 GCS resources traversed, 0 cancelled
0 GCS shadows traversed, 0 replayed, 0 duplicates
lms 0 finished replaying gcs resources
0 write requests issued in 0 GCS resources
0 PIs marked suspect, 0 flush PI msgs
0 write requests issued in 0 GCS resources
0 PIs marked suspect, 0 flush PI msgs
0 write requests issued in 0 GCS resources
0 PIs marked suspect, 0 flush PI msgs
0 write requests issued in 0 GCS resources
0 PIs marked suspect, 0 flush PI msgs
0 write requests issued in 0 GCS resources
0 PIs marked suspect, 0 flush PI msgs
lms 0 finished fixing gcs write protocol
*** 2015-11-12 15:46:40.320
0 GCS shadows cancelled, 0 closed
0 GCS resources traversed, 0 cancelled
0 GCS shadows traversed, 0 replayed, 0 duplicates
lms 0 finished replaying gcs resources
0 write requests issued in 0 GCS resources
0 PIs marked suspect, 0 flush PI msgs
0 write requests issued in 0 GCS resources
0 PIs marked suspect, 0 flush PI msgs
0 write requests issued in 0 GCS resources
0 PIs marked suspect, 0 flush PI msgs
0 write requests issued in 0 GCS resources
0 PIs marked suspect, 0 flush PI msgs
0 write requests issued in 0 GCS resources
0 PIs marked suspect, 0 flush PI msgs
lms 0 finished fixing gcs write protocol
kjbmpsyncack: domain 0, from 1, type = 1
kjbmpsyncack[4,0]: complete
跟蹤下LMD
SQL> oradebug setospid 5181
Oracle pid: 6, Unix process pid: 5181, image: oracle@jingfa1 (LMD0)
SQL> oradebug event 10708 trace name context forever,level 12;
Statement processed.
SQL> insert into t_boc select level from dual connect by level<=10000;
10000 rows created.
SQL> commit;
Commit complete.
可見如果之前手工刪除LMD的跟蹤TRC檔案，再次生成是空的檔案或生成的檔案不存在
SQL> oradebug event 10708 trace name context off;
Statement processed.
SQL> oradebug tracefile_name
/u01/app/oracle/admin/jingfa/bdump/jingfa1_lmd0_5181.trc
SQL> oradebug setospid 5181
Oracle pid: 6, Unix process pid: 5181, image: oracle@jingfa1 (LMD0)
SQL> oradebug event 10708 trace name context forever,level 12;
Statement processed.
SQL> oradebug flush
Statement processed.
SQL> oradebug flush
Statement processed.
SQL> oradebug flush
Statement processed.
SQL> oradebug flush
Statement processed.
SQL> oradebug event 10708 trace name context off;
Statement processed.
SQL> oradebug tracefile_name
/u01/app/oracle/admin/jingfa/bdump/jingfa1_lmd0_5181.trc
SQL> host ll -l /u01/app/oracle/admin/jingfa/bdump/jingfa1_lmd0_5181.trc
/bin/bash: ll: command not found
SQL> host ls -l /u01/app/oracle/admin/jingfa/bdump/jingfa1_lmd0_5181.trc
-rw-r--r-- 1 oracle oinstall 0 Nov 12 17:52 /u01/app/oracle/admin/jingfa/bdump/jingfa1_lmd0_5181.trc
SQL> oradebug setmypid
Statement processed.
SQL> oradebug dump events 1
Statement processed.
SQL> oradebug tracefile_name
/u01/app/oracle/admin/jingfa/udump/jingfa1_ora_6610.trc
SQL> host ls -l /u01/app/oracle/admin/jingfa/udump/jingfa1_ora_6610.trc
-rw-r----- 1 oracle oinstall 752 Nov 12 18:21 /u01/app/oracle/admin/jingfa/udump/jingfa1_ora_6610.trc
可見資訊非常少，也不懂為何義
Dump event group for level SESSION
TC Addr Evt#(b10) Action TR Addr Arm Life
加大級別
SQL> oradebug setmypid
Statement processed.
SQL> oradebug dump events 12
Statement processed.
SQL> oradebug tracefile_name
/u01/app/oracle/admin/jingfa/udump/jingfa1_ora_6610.trc
提示級別不對，有效的級別為：會話是1，程式是2，系統是4
Bad group level (12) passed to dump routine
Valid values are (1) Session (2) Process (4) System)
轉儲程式
SQL> oradebug setmypid
Statement processed.
SQL> oradebug dump events 2
Statement processed.
SQL> oradebug tracefile_name
/u01/app/oracle/admin/jingfa/udump/jingfa1_ora_6610.trc
*** 2015-11-12 18:25:34.673
Dump event group for level PROCESS
TC Addr Evt#(b10) Action TR Addr Arm Life
轉儲系統
SQL> oradebug setmypid
Statement processed.
SQL> oradebug dump events 4
Statement processed.
SQL> oradebug tracefile_name
/u01/app/oracle/admin/jingfa/udump/jingfa1_ora_6610.trc
Dump event group for level SYSTEM
TC Addr Evt#(b10) Action TR Addr Arm Life
不過上述全沒有資訊，僅是個條目
SQL> oradebug setmypid
Statement processed.
SQL> oradebug dump events 4
Statement processed.
SQL> oradebug flush
Statement processed.
SQL> oradebug flush
Statement processed.
SQL> oradebug flush
Statement processed.
SQL> oradebug tracefile_name
/u01/app/oracle/admin/jingfa/udump/jingfa1_ora_6610.trc
為何僅有條目，沒有內容呢？肯定是少執行或配置了什麼命令
我們再好好分析下TRC檔案產生的條目，好像是要和跟蹤的EVENT有關，你看evt#我估計就是event的編號，arm我估計就是對應EVENT的報錯資訊
TC Addr Evt#(b10) Action TR Addr Arm Life
我們在當前會話跟蹤一個4031事件
SQL> alter session set events '4031 trace name errorstack forever,level 12';
Session altered.
SQL> oradebug setmypid
Statement processed.
SQL> oradebug dump events 1
Statement processed.
SQL> oradebug tracefile_name
/u01/app/oracle/admin/jingfa/udump/jingfa1_ora_6610.trc
看到沒，這下有資料了
Dump event group for level SESSION
TC Addr Evt#(b10) Action TR Addr Arm Life
2F3DED08 4031 1 2f3d00f8 0 0
TR Name TR level TR address TR arm TR life TR type
ERRORSTACK 12 0 -1 2 0
關閉了事件，TRC檔案又只有條目資訊了
SQL> alter session set events '4031 trace name errorstack off';
Session altered.
SQL> oradebug setmypid
Statement processed.
SQL> oradebug dump events 1
Statement processed.
SQL> oradebug tracefile_name
/u01/app/oracle/admin/jingfa/udump/jingfa1_ora_6610.trc
*** 2015-11-12 18:36:25.257
Dump event group for level SESSION
TC Addr Evt#(b10) Action TR Addr Arm Life
好，我們繼續分析10708 event,可見對於oracebug event,即使配置一個不存在的event,oradebug也會配置成功，不會報任何的錯誤
SQL> oradebug setmypid
Statement processed.
SQL> oradebug event 888888888 trace name context forever,level 12;
Statement processed.
可以直接檢視event的含義，採用oerr ora event
SQL> host oerr ora 10708
10708, 00000, "print out trace information from the RAC buffer cache"
// *Cause: N/A
// *Action: THIS IS NOT A USER ERROR NUMBER/MESSAGE. THIS DOES NOT NEED TO BE
// TRANSLATED OR DOCUMENTED. IT IS USED ONLY FOR DEBUGGING.
SQL> host oerr ora 888888888
不用oracebug,換成alter session set events,但是內容還是空的
SQL> alter session set tracefile_identifier='hbstest';
Session altered.
SQL> alter session set events '10708 trace name context forever,level 12';
Session altered.
SQL> alter session set events '10708 trace name context off';
Session altered.
繼續改用alter system set events,仍然沒有內容
SQL> alter session set tracefile_identifier='hbstest';
Session altered.
SQL> alter system set events '10708 trace name context forever,level 12';
System altered.
SQL> create table t_buffer_dml(a int);
Table created.
SQL> insert into t_buffer_dml values(1);
1 row created.
SQL> commit;
Commit complete.
SQL> alter session set events '10708 trace name context off';
Session altered.
再換個思路
SQL> alter system set event='10708 trace name context forever,level 10' scope=spfile;
System altered.
SQL> create pfile='/home/oracle/10gpfile.txt' from spfile;
File created.
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
SQL> startup
ORACLE instance started.
Total System Global Area 599785472 bytes
Fixed Size 2022632 bytes
Variable Size 335545112 bytes
Database Buffers 260046848 bytes
Redo Buffers 2170880 bytes
Database mounted.
Database opened.
[oracle@jingfa1 udump]$ ll -lrht
-rw-r----- 1 oracle oinstall 65 Nov 12 19:10 jingfa1_ora_19482_hbstest.trc
-rw-r----- 1 oracle oinstall 65 Nov 12 19:12 jingfa1_ora_26927_hbstest.trc
-rw-r----- 1 oracle oinstall 1.5K Nov 12 19:37 jingfa1_ora_31562.trc
-rw-r----- 1 oracle oinstall 969 Nov 12 19:38 jingfa1_ora_1779.trc
-rw-rw---- 1 oracle oinstall 593 Nov 12 19:38 jingfa1_ora_2035.trc
-rw-r----- 1 oracle oinstall 759 Nov 12 19:38 jingfa1_ora_1981.trc
-rw-rw---- 1 oracle oinstall 593 Nov 12 19:38 jingfa1_ora_2234.trc
-rw-rw---- 1 oracle oinstall 593 Nov 12 19:38 jingfa1_ora_2233.trc
-rw-r----- 1 oracle oinstall 37K Nov 12 19:38 jingfa1_ora_2261.trc
-rw-rw---- 1 oracle oinstall 3.2K Nov 12 19:39 jingfa1_ora_2224.trc
[oracle@jingfa1 udump]$ date
kclmap: mapping region: 0x60400000 length: 595591168
*** 2015-11-12 19:39:39.982
*** SERVICE NAME:(SYS$USERS) 2015-11-12 19:39:39.982
*** SESSION ID:(146.1) 2015-11-12 19:39:39.981
kclscrs: req=0 block=1/400760
kclscrs: req=0 typ=immediate mode=S
kclscrs: req=0 block=1/400761
kclscrs: req=0 typ=immediate mode=S
kclscrs: req=0 block=1/400763
kclscrs: req=0 typ=immediate mode=S
中間略
kclscrs: req=0 block=1/4022ea
kclwcrs: req=0 typ=cur wtyp=2hop tm=749
kclscrs: req=0 block=3/c01633
kclwcrs: req=0 typ=stale
kclscrs: req=0 block=3/c016fc
kclwcrs: req=0 typ=cur wtyp=2hop tm=870
[oracle@jingfa1 udump]$
---在139會話大量產生DML
SQL> select sid from v$mystat where rownum=1;
SID
----------
139
SQL> select paddr from v$session where sid=139;
PADDR
----------------
0000000083A5E580
SQL> select spid from v$process where addr='0000000083A5E580';
SPID
------------
13992
SQL> insert into t_boc select level from dual connect by level<=10000;
10000 rows created.
SQL> insert into t_boc select level from dual connect by level<=10000;
10000 rows created.
SQL> insert into t_boc select level from dual connect by level<=10000;
10000 rows created.
SQL> insert into t_boc select level from dual connect by level<=10000;
10000 rows created.
SQL> commit;
Commit complete.
-rw-r----- 1 oracle oinstall 1.5K Nov 12 19:37 jingfa1_ora_31562.trc
-rw-r----- 1 oracle oinstall 969 Nov 12 19:38 jingfa1_ora_1779.trc
-rw-rw---- 1 oracle oinstall 593 Nov 12 19:38 jingfa1_ora_2035.trc
-rw-r----- 1 oracle oinstall 759 Nov 12 19:38 jingfa1_ora_1981.trc
-rw-rw---- 1 oracle oinstall 593 Nov 12 19:38 jingfa1_ora_2234.trc
-rw-rw---- 1 oracle oinstall 593 Nov 12 19:38 jingfa1_ora_2233.trc
-rw-r----- 1 oracle oinstall 37K Nov 12 19:38 jingfa1_ora_2261.trc
-rw-rw---- 1 oracle oinstall 3.2K Nov 12 19:39 jingfa1_ora_2224.trc
-rw-r----- 1 oracle oinstall 12K Nov 12 19:46 jingfa1_ora_13992.trc
檢視下jingfa1_ora_13992.trc內容，這個其實就是global buffer cache的內容，可見上述的檔案是基於不同的會話進行分別建立的，
另外：alter set events要生效，必須重啟資料庫，哈哈，終於10708event有值了
kcllkop: req=12 new block=5/1406a97
kcllkop: req=13 new block=5/1406a98
kcllkop: req=0 typ=grant wtyp=2hop tm=1378
kcllkop: req=1 typ=grant wtyp=2hop tm=1380
kcllkop: req=2 typ=grant wtyp=2hop tm=1380
kcllkop: req=3 typ=grant wtyp=2hop tm=1378
kcllkop: req=4 typ=grant wtyp=2hop tm=1377
kcllkop: req=5 typ=grant wtyp=2hop tm=1375
kcllkop: req=6 typ=grant wtyp=2hop tm=1374
kcllkop: req=7 typ=grant wtyp=2hop tm=1373
kcllkop: req=8 typ=grant wtyp=2hop tm=1372
kcllkop: req=9 typ=grant wtyp=2hop tm=1371
kcllkop: req=10 typ=grant wtyp=2hop tm=1369
kcllkop: req=11 typ=grant wtyp=2hop tm=1369
kcllkop: req=12 typ=grant wtyp=2hop tm=1368
kcllkop: req=13 typ=grant wtyp=2hop tm=1365
來自 “ ITPUB部落格 ” ，連結：http://blog.itpub.net/9240380/viewspace-1834540/，如需轉載，請註明出處，否則將追究法律責任。
相關文章
- ORACLE EVENT && ORADEBUGOracle
- （轉）Oracle EVENT && ORADEBUGOracle
- 基於oracle 10.2.0.1 rac使用oradebug dump hanganalyze 分析oracle hang系列四Oracle
- 基於oracle 10.2.0.1 rac使用oradebug dump hanganalyze 分析oracle hang系列五Oracle
- 基於oracle 10.2.0.1 rac使用oradebug dump hanganalyze 分析oracle hang系列六Oracle
- 基於oracle 10.2.0.1 rac學習lms程式系列四Oracle
- 使用oradebug dump hanganalyze分析oracle hang系列一Oracle
- 使用oradebug dump hanganalyze 分析oracle hang系列二Oracle
- 使用oradebug dump hanganalyze 分析oracle hang系列三Oracle
- 在oracle10g及oracle11g使用oradebug生成systemstate dump檔案系列二Oracle
- ORACLE 配置event引數Oracle
- oracle buffer cache管理機制_buffer cache dump與lru機制小記Oracle
- 由oradebug poke process allocation latch引發dfs lock handle等待事件進一步分析事件
- 常用的DUMP語句ORADEBUG語法
- Oracle expdp impdp dump引數介紹Oracle
- 通過oracle event來dump資料檔案頭資訊Oracle
- Oracle資料庫event事件與dump檔案介紹Oracle資料庫事件
- wait event:gc buffer busyAIGC
- 由oradebug poke推進scn理解scn base及scn wrap系列一
- oracle event heapdump_sga dump_shared pool_pga_ugaOracle
- 由吃飯引發的思考
- oracle dg後臺程式及wait event--轉載-精OracleAI
- 由節流函式引發出我對event-loop的思考，順便刷刷爆款題函式OOP
- log buffer space wait event等待事件AI事件
- Oracle oradebug命令Oracle
- 【RAC】Oracle叢集重要程式evmd、cssd和crsd發源地及跟蹤探索OracleCSS
- 由一個emoji引發的思考
- 由春節回家難引發的思考
- oracle rac 核心引數詳解Oracle
- Oracle RAC引數檔案管理Oracle
- ORACLE RAC SGA引數調整Oracle
- 由京東發貨引發的思考和分析
- Oracle RAC一鍵部署004(RAC引數校驗)Oracle
- ORACLE RAC GUARD配置引數——RAC GUARD概念和管理Oracle
- 【RAC】Oracle rac 如何修改公網及vipOracle
- Oracle事件跟蹤及結構資料dumpOracle事件
- lms程式耗用大量記憶體記憶體
- oradebug分析oracle hangOracle