tail -f in windows

…sau la ce poți folosi tool-ul Automatic Diagnostic Repository Command Interpreter. Pentru cei care nu au folosit acest tool, menționez ca a apărut o data cu versiunea 11g R1 cu scopul de a simplifica următoarele taskuri:
– vizualizarea alertlog-ului și incidentelor
– generarea pachetelor zip pentru a fi trimise către suportul Oracle
– managementul fișierelor trace

Prin acest articol doresc să fac un rezumat al comenzile esențiale, pe care un DBA trebuie să le știe, în scopul de a folosi adrci.

1. Vizualizarea alertlog-ului

Executabilul poate fi apelat din directorul $ORACLE_HOME\bin

>adrci

ADRCI: Release 11.1.0.7.0 - Production on Tue Sep 18 11:20:34 2012

Copyright (c) 1982, 2007, Oracle.  All rights reserved.

ADR base = "f:\oracle"
adrci>

In funcție de infrastructura, puteți avea mai multe adrese ADR Home.

adrci> show home
ADR Homes:
diag\clients\user_system\host_2488280724_11
diag\clients\user_unknown\host_2488280724_11
diag\clients\user_unknown\host_3642107604_11
diag\rdbms\prod\prod
diag\tnslsnr\listener

Pentru a verifica alertlog-ul, setăm mai întâi home-ul bazei de date și folosim comanda show alert cu opțiunea -tail -f

adrci> set home diag\rdbms\prod\prod
adrci> show alert -tail -f
2012-09-18 01:00:13.840000 +03:00
  Current log# 3 seq# 18660 mem# 0: F:\ORACLE\ORADATA\PROD\REDO03A.LOG
  Current log# 3 seq# 18660 mem# 1: F:\ORACLE\ORADATA\PROD\REDO03B.LOG
Thread 1 advanced to log sequence 18661 (LGWR switch)
  Current log# 4 seq# 18661 mem# 0: F:\ORACLE\ORADATA\PROD\REDO04A.LOG
  Current log# 4 seq# 18661 mem# 1: F:\ORACLE\ORADATA\PROD\REDO04B.LOG
2012-09-18 01:00:14.996000 +03:00
LNS: Standby redo logfile selected for thread 1 sequence 18661 for destination LOG_ARCHIVE_DEST_2
2012-09-18 02:00:00.099000 +03:00
Clearing Resource Manager plan via parameter
2012-09-18 05:00:14.157000 +03:00
Thread 1 advanced to log sequence 18662 (LGWR switch)
  Current log# 1 seq# 18662 mem# 0: F:\ORACLE\ORADATA\PROD\REDO01A.LOG
  Current log# 1 seq# 18662 mem# 1: F:\ORACLE\ORADATA\PROD\REDO01B.LOG
LNS: Standby redo logfile selected for thread 1 sequence 18662 for destination LOG_ARCHIVE_DEST_2
2012-09-18 10:38:53.486000 +03:00
Thread 1 advanced to log sequence 18663 (LGWR switch)
  Current log# 2 seq# 18663 mem# 0: F:\ORACLE\ORADATA\PROD\REDO02A.LOG
  Current log# 2 seq# 18663 mem# 1: F:\ORACLE\ORADATA\PROD\REDO02B.LOG
2012-09-18 10:38:55.267000 +03:00
LNS: Standby redo logfile selected for thread 1 sequence 18663 for destination LOG_ARCHIVE_DEST_2

2. Legătura problemelor și a incidentelor

Un incident este apariția concretă a unei probleme dar o problema poate avea mai multe incidente.
Pentru a vizualiza legătura dintre probleme și incidente urmăriți exemplul de mai jos:

adrci> show problem

ADR Home = f:\oracle\diag\rdbms\prod\prod:
*************************************************************************
PROBLEM_ID           PROBLEM_KEY                                                 LAST_INCIDENT
-------------------- ----------------------------------------------------------- ----------------
19                   ORA 7445 [opiexe()+21950]                                   866138

adrci> show incident

ADR Home = f:\oracle\diag\rdbms\prod\prod:
*************************************************************************
INCIDENT_ID          PROBLEM_KEY                                                 CREATE_TIME
-------------------- ----------------------------------------------------------- ----------------
866138               ORA 7445 [opiexe()+21950]                                   2012-08-07 12:30

adrci> show incident -mode detail -p "incident_id=866138"

ADR Home = f:\oracle\diag\rdbms\prod\prod:
*************************************************************************

**********************************************************
INCIDENT INFO RECORD 1
**********************************************************
   INCIDENT_ID                   866138
   STATUS                        ready
   CREATE_TIME                   2012-08-07 12:30:14.697000 +03:00
   PROBLEM_ID                    19
   CLOSE_TIME                    
   FLOOD_CONTROLLED              none
   ERROR_FACILITY                ORA
   ERROR_NUMBER                  7445
   ERROR_ARG1                    opiexe()+21950
   ERROR_ARG2                    ACCESS_VIOLATION
   ERROR_ARG3                    ADDR:0x58
   ERROR_ARG4                    PC:0x2965A2C
   ERROR_ARG5                    UNABLE_TO_READ
   ERROR_ARG6                    
   ERROR_ARG7                    
   ERROR_ARG8                    
   SIGNALLING_COMPONENT          
   SIGNALLING_SUBCOMPONENT       
   SUSPECT_COMPONENT             
   SUSPECT_SUBCOMPONENT          
   ECID                          
   IMPACTS                       0
   PROBLEM_KEY                   ORA 7445 [opiexe()+21950]
   FIRST_INCIDENT                720831
   FIRSTINC_TIME                 2012-01-03 23:17:27.448000 +02:00
   LAST_INCIDENT                 866138
   LASTINC_TIME                  2012-08-07 12:30:14.697000 +03:00
   IMPACT1                       0
   IMPACT2                       0
   IMPACT3                       0
   IMPACT4                       0
   KEY_NAME                      SID
   KEY_VALUE                     1384.34804
   KEY_NAME                      Client ProcId
   KEY_VALUE                     ORACLE.EXE.4036_6004
   KEY_NAME                      ProcId
   KEY_VALUE                     254.209
   OWNER_ID                      1
   INCIDENT_FILE                 f:\oracle\diag\rdbms\prod\prod\trace\prod_ora_6004.trc
   OWNER_ID                      1
   INCIDENT_FILE                 f:\oracle\diag\rdbms\prod\prod\incident\incdir_866138\prod_ora_6004_i866138.trc
1 rows fetched

adrci>

Daca vrem sa verificam fișierul trace:

adrci> show trace f:\oracle\diag\rdbms\prod\prod\incident\incdir_866138\prod_ora_6004_i866138.trc

3. Crearea fișierelor pachet pentru a fi trimise către Oracle Suport

Dacă nu putem să rezolve singuri o problema atunci o putem trimite către Oracle Suport. Pentru acest lucru adunam toate informațiile necesare cu o metoda numita „Incident Packaging Service” (IPS):

adrci> ips create package problem 19 correlate all
 Created package 4 based on problem id 19, correlation level all
adrci> ips generate package 2 in "f:\oracle"  
Generated package 4 in file f:\oracle\ORA7445qc_20120918.zip, mode complete

4. Managementul Fișierelor trace

Probabil ca ați observat ca Oracle 11g generează o grămadă de fișiere trace, care trebuie curățate din când în când.
De fapt ele sunt curățate automat conform politicii implicite de curățare, dar dacă dorim sa modificam aceasta politica vom proceda în următorul mod

adrci> show tracefile -rt
   23-JUN-12 10:21:34  diag\rdbms\prod\prod\incident\incdir_816660\prod_j002_4952_i816660.trc
   23-JUN-12 06:36:25  diag\rdbms\prod\prod\incident\incdir_816836\prod_j004_4024_i816836.trc
   20-JUN-12 22:21:05  diag\rdbms\prod\prod\incident\incdir_816740\prod_j002_2320_i816740.trc
   20-JUN-12 14:05:26  diag\rdbms\prod\prod\incident\incdir_817892\prod_ora_4112_i817892.trc
   20-JUN-12 13:55:40  diag\rdbms\prod\prod\incident\incdir_817924\prod_ora_2572_i817924.trc
   20-JUN-12 12:11:37  diag\rdbms\prod\prod\incident\incdir_817684\prod_ora_880_i817684.trc
   30-MAR-12 20:17:54  diag\rdbms\prod\prod\incident\incdir_768739\prod_ora_4812_i768739.trc
   15-FEB-12 16:48:17  diag\rdbms\prod\prod\incident\incdir_739792\prod_ora_3640_i739792.trc
   15-FEB-12 16:32:11  diag\rdbms\prod\prod\incident\incdir_737168\prod_ora_4792_i737168.trc
   18-JAN-12 17:03:34  diag\rdbms\prod\prod\incident\incdir_737632\prod_ora_1680_i737632.trc
   04-JAN-12 00:17:58  diag\rdbms\prod\prod\incident\incdir_720831\prod_ora_5032_i720831.trc
   17-DEC-11 07:22:04  diag\rdbms\prod\prod\incident\incdir_704752\prod_j002_1812_i704752.trc
   17-DEC-11 07:22:04  diag\rdbms\prod\prod\incident\incdir_704752\prod_m001_4880_i704752_a.tr
   29-NOV-11 23:01:16  diag\rdbms\prod\prod\incident\incdir_704917\prod_j002_3476_i704917.trc
   29-NOV-11 23:00:34  diag\rdbms\prod\prod\incident\incdir_704916\prod_j002_3476_i704916.trc
   28-NOV-11 23:01:32  diag\rdbms\prod\prod\incident\incdir_704897\prod_j000_5640_i704897.trc
   28-NOV-11 23:00:36  diag\rdbms\prod\prod\incident\incdir_704896\prod_j000_5640_i704896.trc
   27-NOV-11 07:00:32  diag\rdbms\prod\prod\incident\incdir_704553\prod_j005_3708_i704553.trc
   27-NOV-11 07:00:10  diag\rdbms\prod\prod\incident\incdir_704552\prod_j005_3708_i704552.trc

adrci> show control

ADR Home = f:\oracle\diag\rdbms\prod\prod:
*************************************************************************
ADRID      SHORTP_POLICY  LONGP_POLICY LAST_MOD_TIME  LAST_AUTOPRG_TIME  LAST_MANUPRG_TIME   
----------------- -------------------- -------------- -------------------------------   ------------------------------------ -----
699077231  720            8760         2010-04-28     19:25:52.435000 +03:00  2012-09-15 03:11:57.485000 +03:00 
1 rows fetched

In mod implicit fișierele trace se vor șterge după 30 de zile (720 de ore) iar incidentele după un an (8760 de ore). Vom modifica această politică după cum urmează:

adrci> set control (SHORTP_POLICY = 180)

adrci> set control (LONGP_POLICY = 1460)

Dacă dorim să ștergem manual fisierele trace mai vechi de o zi (1440 de minute) atunci executam:

adrci> purge -age 1440 -type trace

Pentru mai multe informații despre acest tool consultați manualul:
http://docs.oracle.com/cd/B28359_01/server.111/b28319/adrci.htm