Category: Tools

Errors and Bugs, Tools

ORA-19751: could not create the change tracking file

 , , , ,

Let’s make it simple to solve the problem:

# Error:

SQL> alter database open;
alter database open
*
ERROR at line 1:
ORA-19751: could not create the change tracking file
ORA-19750: change tracking file: '+DGDATA/mydb/changetracking/ctf.470.859997781'
ORA-17502: ksfdcre:1 Failed to create file +DGDATA/mydb/changetracking/ctf.470.859997781
ORA-17501: logical block size 4294967295 is invalid
ORA-15001: diskgroup "DGDATA" does not exist or is not mounted
ORA-17503: ksfdopn:2 Failed to open file +DGDATA/mydb/changetracking/ctf.470.859997781
ORA-15001: diskgroup "DGDATA" does not exist or is not mounted
ORA-15001: diskgroup "DGDATA" does not exist or is not mounted

# Solution:

SQL> alter database disable BLOCK CHANGE TRACKING;
Database altered.
SQL> alter database open;
Database altered.

Then, after everything be OK, you fix the situation recrating a BCTF:

ALTER DATABASE ENABLE BLOCK CHANGE TRACKING USING FILE '+DGDATANEW';

MTFBWU!

Matheus.

Backup and Restore, Database, Migration | Upgrades | Patching, Tools

Database Migration/Move with RMAN: Are you sure nothing is missing?

 , ,

Forced by the destiny to make a migration using backup/restore (with a little downtime), how to be sure nothing will be lost during the migration?
Here is a way: Create your own data just before migrating. 🙂

Seems like a kludge and it is.. haha.. But it works. Take a look:

# Original Database

SQL> shu immediate;
Database closed.
Database dismounted.
ORACLE instance shut down.
SQL> startup restrict;
ORACLE instance started.
Total System Global Area 2689060864 bytes
Fixed Size 2229520 bytes
Variable Size 1996491504 bytes
Database Buffers 671088640 bytes
Redo Buffers 19251200 bytes
Database mounted.
Database opened.
SQL> create table matheus_boesing.migration (text varchar2(10));
Table created.
SQL> insert into matheus_boesing.migration values ('well done!');
1 row created.
SQL> commit;
Commit complete.
SQL> alter system switch logfile;
System altered.
SQL> /
System altered.
SQL> /
System altered.
SQL> /
System altered.
SQL> /
System altered.
SQL> /
System altered.
SQL> shu immediate;
SQL> exit;
$ rman target /
connect catalog rman_mydb/password@catalogdb
run { backup archivelog all;}

# Destination Database

$ rman target /
connect catalog rman_mydb/password@catalogdb
run { recover database;}
$ sqlplus / as sysdba
SQL> select count(1), to_char(CHECKPOINT_TIME, 'DD/MM/YYYY HH24:MI:SS') from V$DATAFILE_HEADER t
group by to_char(CHECKPOINT_TIME, 'DD/MM/YYYY HH24:MI:SS') order by 2;
COUNT(1) TO_CHAR(CHECKPOINT_
---------- -------------------
51 27/06/2015 22:15:28
-- All datafiles with synchronized headers...
SQL> alter database open read only; 
-- If needed, you can do more recover, this way...
Database altered.
SQL> select * from matheus_boesing.migration;
TEXT
----------
well done!
-- Means no more recover is needed :)
SQL> shutdown immediate;
SQL> alter database open resetlogs;
Database altered.

And be Happy! 😀

Matheus.

Database, Linux, Shell Scripts, Tools

Grepping Entries from Alert.log

 , , , , , ,

Hey hey,
One more McGayver by me! Haha
Again to find some information in alert. This time, I’m looking to count and list all occurrences of an action in alert. To archive this, I made the script below.

grep-swiss-knife-590x295

The functionality is just a little bit more complex than the script of the last post, but stills quite simple. Take a look:

Parameters:
PAR1: name of alert (the main alert.log)
PAR2: Searched token
PAR3: Start day you want to, in the format “Mon dd” or just “Mon”. Below an example.
PAR4: Start Year (4 digits)
PAR5: [optional]End day you want to, in the format “Mon dd” or just “Mon”. The default value is “until now”.
PAR6: [optional]End Year (4 digits). The default value is “until now”. If you use the PAR5, you have to use PAR6.
PAR7: [optional] List All entries and when?. If you want to use this PAR, you must to use PAR5 and PAR6.

Examples (Looking for service reconfigurations):
Ex1: sh grep_entries_alert.sh alert_xxdb_1.log “services=” “Apr 12” 2015
(Seach between April 12 and now and count entries).
Ex2: sh grep_entries_alert.sh alert_xxdb_1.log “services=” “Apr 01” 2015 “May 30” 2015
(Seach between April 01 and May 30 and count the entries).
Ex3: sh grep_entries_alert.sh alert_xxdb_1.log “services=” “Apr 01” 2015 “May 30” 2015 LIST
(Seach between April 01 and May 30 and count the entries and list them all…)

# Script grep_entries_alert.sh
if [ $# -lt 6 ]; then
FIN=`cat $1 |wc -l`
else FIN=`cat $1 |grep -n $5 |grep $6$ |head -n 1 |cut -d':' -f1`
fi
BEG=`cat $1 |grep -n "$3" |grep $4$ |head -n 1 |cut -d':' -f1`
NMB=`expr $FIN - $BEG`
ENTR=`cat $1 |head -n $FIN |tail -$NMB| grep $2|wc -l`
echo Number of Entries: $ENTR >log.log
if [ $# -lt 7 ]; then
echo ------- Complete List Of Entries and When ---------- >> log.log
for line in `cat $1 |head -n $FIN |tail -$NMB| grep -n $2|cut -d':' -f1`;do
LR=`expr $line + $BEG` # To get "real line", without the displacement
DAT=`expr $LR - 1`     # To get line date of entry
echo awk \'NR==$DAT\' $1 >>aux.sh # Printing the lines just calculted
echo awk \'NR==$LR\' $1 >>aux.sh  # with aux.sh
done;
sh aux.sh >>log.log
fi
cat log.log

It’s not beautiful. But it works! 🙂

After that, there is the new blog sponsor:
MacGyver-macgyver-880400_200_228
(Hahahaha)

Matheus.

Database, Linux, Shell Scripts, Tools

Grepping Alert by Day

 , , , , , ,

Hi all,
For that moment when your alert is very big and some OS doesn’t “work very well with it” (in my case was using AIX), I jerry-ringged the shellscript bellow. It puts in a new log just the log entries of a selected day.

24 7 365

The call can be made with two or three parameters, this way:

Parameters:
PAR1: name of alert (the main alert.log)
PAR2: Day you want to, in the format “Mon dd”. Below an example.
PAR3: [optional] desired year. The default is the current year. But is useful specially on the “new year” period…

Examples:
Ex1: sh grep_day.sh alert_xxdb_1.log “Apr 12”
Ex2: sh grep_day.sh alert_xxdb_1.log “Apr 12” 2014

Generated files:
dalert_2015Apr12.log
dalert_2014Apr12.log

# Script grep_day.sh
if [ $# -lt 3 ]; then
YEAR=`date +"%Y"`
else YEAR=$3
fi
DATEFORMAT=`echo $2|cut -d' ' –f1`""`echo $2|cut -d' ' –f2`
BEG=`cat $1 |grep -n "$2" |grep $YEAR |head -1 |cut -d':' -f1`
FIN=`cat $1 |grep -n "$2" | grep $YEAR |tail -1 |cut -d':' -f1`
NMB=`expr $FIN - $BEG`
cat $1 |head -$FIN |tail -$NMB > dalert_$YEAR$DATEFORMAT.log

Belive me! It can be useful…. haha

See ya!

Matheus.