Wednesday 17 October 2018

SQL for RAC db

RAC sql RAC Instances, DB instances, instance status, List of all instances, with startup time, up till, host load. col host_name for a16 select i.instance_number, i.host_name, i.instance_name, to_char(startup_time, 'DD Mon YYYY HH24:Mi') startup_time, floor(sysdate - startup_time) uptill, i.status, round(o.value, 3) value from gv$instance i, gv$osstat o where i.instance_number = o.inst_id and o.stat_name = 'LOAD' order by instance_number; -- e.g. INSTANCE_NUMBER HOST_NAME INSTANCE_NAME STARTUP_TIME UPTILL STATUS VALUE --------------- ---------------- ---------------- -------------------------- ---------- ------------ ---------- 1 proddbx00101 dex011 12 Feb 2018 02:52 247 OPEN 15.979 2 proddbx00102 dex012 12 Feb 2018 05:25 247 OPEN 18 3 proddbx00103 dex013 12 Feb 2018 07:23 247 OPEN 21.59 4 proddbx00104 dex014 12 Feb 2018 09:37 247 OPEN 19.079

Current Transaction for RAC set line 250 col "username_sid_inst" for a35 select s.username||'('||s.sid||', '||s.inst_id||')' username_sid_inst,s.sid,t.start_time, t.used_ublk as "Block Count", round(t.used_ublk*8192/1024/1024,2) Size_MB from gv$transaction t,gv$session s where s.saddr=t.ses_Addr and t.inst_id=s.inst_id order by start_time;

Session Wait for RAC set line 250 col username for a24 col event for a45 col 'FileId# Obj#' for a16 select s.inst_id, s.username, s.row_wait_file# ||', '|| s.ROW_WAIT_OBJ# as "FileId# Obj#", s.sql_hash_value, sw.event, sw.wait_time, sw.seconds_in_wait, sw.state from gv$session s, gv$session_wait sw where s.sid = sw.sid and s.inst_id = sw.inst_id and sw.wait_time != 0 order by s.inst_id, s.username;

Sessions connected for RAC -- Group by User (Summary) select s.username, count(*) from gv$session s group by s.username order by s.username; -- Group by Instance, User select s.inst_id, s.username, count(*) from gv$session s group by s.inst_id, s.username order by s.inst_id, s.username;

Locks (Holder/Waiter) for RAC column sess format A20 select substr(DECODE(request,0,'Holder: ','Waiter: ')||sid||' ('||inst_id||')',1,18) sess, id1, id2, lmode, request, type, to_char(to_date(ctime,'sssss'),' hh24:mi:ss') time_h_m_s from gv$lock where (id1, id2, type) IN (select id1, id2, type from gv$lock where request>0) order by id1, request, ctime desc;

Thursday 4 October 2018

PDB DBA user

create user c##dbareadonly identified by urdbapa55 container=all; grant create session to c##dbateam container=all; grant set container to c##dbateam container=all; grant select_catalog_role to c##dbateam container=all; alter user c##dbateam set container_data=all container = current; Above (last line is important, otherwise dba user can't select * from v$pdbs, can't select from pdb views)

Favorite SQL for session waits

Session waits col "FileId# Obj#" for a18 col event for a42 select s.inst_id, s.username, s.row_wait_file# ||', '|| s.ROW_WAIT_OBJ# as "FileId# Obj#", s.sql_hash_value, sw.event, sw.wait_time, sw.seconds_in_wait, sw.state from gv$session s, gv$session_wait sw where s.sid = sw.sid and s.inst_id = sw.inst_id and sw.wait_time != 0 order by s.inst_id, s.username Other related, or common sql -- Current Transaction col "username_sid" for a35 select s.username||'('||s.sid||')' username_sid,s.sid,t.start_time,t.used_ublk as "Block Count", round(t.used_ublk*8192/1024/1024,2) Size_MB from v$transaction t,v$session s where s.saddr=t.ses_Addr order by start_time; -- Long OPS select sid, serial#, context, sofar, totalwork, target, sql_hash_value, round(sofar/totalwork*100,2) "%_COMPLETE", opname, to_char(start_time, 'DD Mon YYYY - HH24:MI') start_time, round(time_remaining/60) tr, to_char(last_update_time, 'DD Mon YYYY - HH24:MI') last_update_time from v$session_longops where totalwork != 0 and sofar <> totalwork order by start_time; SQL from hash value, SQL hash set line 200 select * from v$sql where hash_value = 9999999; select sq.sql_text from v$sqltext_with_newlines sq where sq.hash_value = 9999999 order by sq.piece SELECT s.username, sl.sid, sq.executions, sl.last_update_time, sl.sql_id, sl.sql_hash_value, opname, target, elapsed_seconds, time_remaining, sq.sql_fulltext FROM v$session_longops sl INNER JOIN v$sql sq ON sq.sql_id = sl.sql_id INNER JOIN v$session s ON sl.SID = s.SID AND sl.serial# = s.serial# WHERE time_remaining > 0 SQL from pid / session from Process Id / Session Id set line 200 col schemaname for a20 col host_name for a20 col program for a16 col pprogram for a20 col machine for a20 col com for a10 -- PID version select p.spid, s.schemaname, s.sid, s.serial#, s.status, s.machine, s.program, p.program pprogram, s.sql_hash_value, s.sql_id, s.prev_sql_id, decode(command, 2, 'INSERT', 3, 'SELECT', 6, 'UPDATE', 7, 'DELETE', 47, 'PL/SQL EXE', command) com , i.host_name, i.instance_name from v$session s, v$process p, v$instance i where s.paddr = p.addr and p.spid = 22844; -- SID version col "Sid_Ser#_SqlHash"for a24 select p.spid, s.schemaname, s.sid||','||s.serial#||' ['||s.sql_hash_value||']' as "Sid_Ser#_SqlHash", s.status, s.machine, s.program, p.program pprogram, s.sql_id, s.prev_sql_id, decode(command, 2, 'INSERT', 3, 'SELECT', 6, 'UPDATE', 7, 'DELETE', 47, 'PL/SQL EXE', command) com , i.host_name, i.instance_name from v$session s, v$process p, v$instance i where s.paddr = p.addr and s.sid = 2790;

Wednesday 12 September 2018

RMAN point in time restore challenge 12c

RMAN point in time pdb restore (using duplicate), gives error due issue with different pdb. Using Oracle 12.1.0.2.0 This is the restore command, I was using: rman auxiliary / catalog rmanuser/password1@CATDB run{ set UNTIL TIME "to_date('2018-07-15 00:00:01','YYYY-MM-DD HH24:MI:SS')" ; DUPLICATE DATABASE 'proddb2' TO 'NI_REST' PLUGGABLE DATABASE CUSTX_PDB; }

Error [1] faced, due to another pdb issue Skipping pluggable database CUST_PDB5 Skipping pluggable database CUST_PDB6 RMAN-00571: =========================================================== RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS =============== RMAN-00571: =========================================================== RMAN-03002: failure of Duplicate Db command at 09/06/2018 16:44:40 RMAN-05501: aborting duplication of target database RMAN-06813: could not translate pluggable database STG_CUSTYPDB

Workaround, not using catalog rman auxiliary / target sys/password2@proddb2 run{ set UNTIL TIME "to_date('2018-07-15 00:00:01','YYYY-MM-DD HH24:MI:SS')" ; DUPLICATE DATABASE TO 'NI_REST' PLUGGABLE DATABASE CUSTX_PDB; }

Error [2] now faced, due to missing arc Executing: alter database datafile 832 offline drop Executing: alter database datafile 833 offline drop Executing: alter database datafile 834 offline drop starting media recovery unable to find archived log archived log thread=1 sequence=13054 Oracle Error: ORA-01547: warning: RECOVER succeeded but OPEN RESETLOGS would get error below ORA-01194: file 3 needs more recovery to be consistent ORA-01110: data file 3: '+DATA/NI_REST/DATAFILE/sysaux.675.986210261' RMAN-00571: =========================================================== RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS =============== RMAN-00571: =========================================================== RMAN-03002: failure of Duplicate Db command at 09/07/2018 13:24:29 RMAN-05501: aborting duplication of target database RMAN-03015: error occurred in stored script Memory Script RMAN-06054: media recovery requesting unknown archived log for thread 1 with sequence 13054 and starting SCN of 7885151161 RMAN>

Workaround
  1. Restored around 15 archive log, starting from above sequence (13054, also for thread 2)
  2. Redone the RMAN clone command
  3. Opening db gave error with control file
  4. Create backup control file, used RESETLOGS block
    1. Removed all non NI_REST datafiles
    2. Kept only 3 LOGFILE, they do not exits on filesystem
  5. Run the backup control (notice has parent db name (PRODDB2) and resetlogs)
  6. Start db with open resetlog
  7. Opened the pdb successfully
  8. Recreated the temp file
With the above, restore was successful. exported the schema from pdb and aux db dropped.


Appendix: catalog start with /u01/oracle/product/12102_01/dbs/' noprompt; alter database backup controlfile to trace as '/home/oracle/mydb_new.ctl.txt'; STARTUP NOMOUNT CREATE CONTROLFILE REUSE DATABASE "PRODDB2" RESETLOGS ARCHIVELOG MAXLOGFILES 192 MAXLOGMEMBERS 3 MAXDATAFILES 1024 MAXINSTANCES 32 MAXLOGHISTORY 584 LOGFILE GROUP 1 '+DATA/NI_REST/ONLINELOG/group_1.262.903320613' SIZE 1024M BLOCKSIZE 512, GROUP 2 '+DATA/NI_REST/ONLINELOG/group_2.263.903320623' SIZE 1024M BLOCKSIZE 512, GROUP 3 '+DATA/NI_REST/ONLINELOG/group_3.269.903321099' SIZE 1024M BLOCKSIZE 512 -- STANDBY LOGFILE DATAFILE '+DATA/NI_REST/DATAFILE/system.693.986210261', '+DATA/NI_REST/DATAFILE/sysaux.675.986210261', '+DATA/NI_REST/DATAFILE/undotbs1.683.986210259', '+DATA/NI_REST/2B4E02B966694E2CE0533F1C6F0A31F5/DATAFILE/system.461.986212971', '+DATA/NI_REST/2B4E02B966694E2CE0533F1C6F0A31F5/DATAFILE/sysaux.467.986212969', '+DATA/NI_REST/DATAFILE/undotbs2.474.986213075', '+DATA/NI_REST/2DB32189A8D41995E0533F1C6F0A911E/DATAFILE/system.459.986212545', '+DATA/NI_REST/2DB32189A8D41995E0533F1C6F0A911E/DATAFILE/sysaux.460.986212545', '+DATA/NI_REST/2DB32189A8D41995E0533F1C6F0A911E/DATAFILE/default_pdb_ts.611.986212545', '+DATA/NI_REST/2DB32189A8D41995E0533F1C6F0A911E/DATAFILE/users.561.986213451' CHARACTER SET AL32UTF8 ; alter database OPEN RESETLOGS; alter tablespace TEMP add tempfie '+DATA' size 150m;

Appendix: error SQL> alter database open resetlogs; alter database open resetlogs * ERROR at line 1: ORA-19838: Cannot use this control file to open database

Thursday 21 June 2018

PDB Tips

List pdbs show pdbs select * from v$pdbs; # All PDB (inc. size) col name for a30 select name, open_mode, total_size/1024/1024 as size_mb from v$pdbs where name like '%'; PDB sessions, pdb connections, pdb sess col name for a20 col username for a32 col machine for a20 select s.inst_id, p.name, s.username, s.machine, count(*) as "SESSION COUNT" from gv$session s, gv$pdbs p where p.con_id = s.con_id and s.inst_id = p.inst_id group by s.inst_id, p.name, s.username, s.machine order by p.name, s.machine; -- List only STG_BIZ_X select s.inst_id, p.name, s.username, s.machine, s.sid, s.serial#, s.status from gv$session s, gv$pdbs p where p.con_id = s.con_id and s.inst_id = p.inst_id and p.name = 'STG_BIZ_X'; INST_ID NAME USERNAME MACHINE SID SERIAL# STATUS ------- ------------- ------------- ------------ --------- ------- -------- 1 STG_BIZ_X SYS ukdevdb017 442 28213 ACTIVE 1 STG_BIZ_X C##JPTADMIN ukdevdb602 777 56244 KILLED 2 STG_BIZ_X C##JPTADMIN ukdevdb601 66 64486 KILLED 2 STG_BIZ_X SYS ukdevdb017 1959 13580 ACTIVE 2 STG_BIZ_X C##JPTADMIN ukdevdb602 1961 46827 KILLED

Grid installation prerequisites

Getting error

[grid@proddev1 ~]$ cd /u01/app/grid/12201 [grid@proddev1 12201]$ ./runcluvfy.sh stage -pre crsinst -n proddev1,proddev2 -fixup -verbose ERROR: PRVG-10467 : The default Oracle Inventory group could not be determined. Verifying '/tmp/' ...FAILED (PRVG-1901) Pre-check for cluster services setup was unsuccessful on all the nodes. Failures were encountered during execution of CVU verification request "stage -pre crsinst". Verifying '/tmp/' ...FAILED PRVG-1901 : failed to setup CVU remote execution framework directory "/tmp/CVU_12.2.0.1.0_grid/" on nodes "proddev1" trprdbr04017: Version of exectask could not be retrieved from node "trprdbr04017" trprdbr04017: Version of exectask could not be retrieved from node "trprdbr04017" CVU operation performed: stage -pre crsinst Date: Jun 21, 2018 9:37:11 AM CVU home: /u01/app/grid/12201/ User: grid


FIX: Correct the fstab, to make sure, its like below exactly

[root@proddev1 ~]# more /etc/fstab |grep -E "/tmp | /opt" /dev/mapper/vg_apol-lv_tmp /tmp ext3 defaults,nodiratime,nodev,nosuid 1 2 /dev/mapper/vg_otc0-lv_opt /opt ext3 defaults,nodiratime,nodev 1 2