Use shell scripts to skillfully collect statistics files

Use shell scripts to skillfully collect statistics files

During data migration, a large number of dump files are generated. You need to perform a simple and clear management of dump files. For example, there are many files in the directory, in addition, some tables are relatively large and correspond to a large number of dump files. In this case, we want to get a very concise report and calculate the number of dump files in each table.

For example, there are 1000 dump files, all of which are based on tables TEST1, TEST2, TEST3, and TEST4. So I want to get a simple statistical report, similar to the following form.

 
 

  
  
TEST1 100  
  
  
TEST2 450  
  
  
TEST3 300  
  
  
TEST4 150 
 
 

Simulate this problem.

Use the following script a. sh to generate the DUMP file.

 
 

  
  
for i in {1..$2}  
  
  
do  
  
  
touch $1_EXT_$i.dmp  
  
  
done 
 
 

Generate the 1000 dump files.

 
 

  
  
[ora11g@rac1 DUMP]$ ksh a.sh TEST1 100  
  
  
[ora11g@rac1 DUMP]$ ksh a.sh TEST2 450  
  
  
[ora11g@rac1 DUMP]$ ksh a.sh TEST3 300  
  
  
[ora11g@rac1 DUMP]$ ksh a.sh TEST4 150  
  
  
[ora11g@rac1 DUMP]$ ll *.dmp|wc -l  
  
  
1000 
 
 

Let's take a look at the dump generated.

 
 

  
  
[ora11g@rac1 DUMP]$ ll *.dmp|tail -10  
  
  
-rw-r--r-- 1 ora11g dba 0 Aug 7 08:13 TEST4_EXT_91.dmp  
  
  
-rw-r--r-- 1 ora11g dba 0 Aug 7 08:13 TEST4_EXT_92.dmp  
  
  
-rw-r--r-- 1 ora11g dba 0 Aug 7 08:13 TEST4_EXT_93.dmp  
  
  
-rw-r--r-- 1 ora11g dba 0 Aug 7 08:13 TEST4_EXT_94.dmp  
  
  
-rw-r--r-- 1 ora11g dba 0 Aug 7 08:13 TEST4_EXT_95.dmp  
  
  
-rw-r--r-- 1 ora11g dba 0 Aug 7 08:13 TEST4_EXT_96.dmp  
  
  
-rw-r--r-- 1 ora11g dba 0 Aug 7 08:13 TEST4_EXT_97.dmp  
  
  
-rw-r--r-- 1 ora11g dba 0 Aug 7 08:13 TEST4_EXT_98.dmp  
  
  
-rw-r--r-- 1 ora11g dba 0 Aug 7 08:13 TEST4_EXT_99.dmp  
  
  
-rw-r--r-- 1 ora11g dba 0 Aug 7 08:13 TEST4_EXT_9.dmp  
  
  
[ora11g@rac1 DUMP]$ 
 
 

First, we need to obtain several related tables in this directory.

For example, in the current situation, you need to obtain

 
 

  
  
TEST1  
  
  
TEST2  
  
  
TEST3  
  
  
TEST4 
 
 

You can use the following command for statistics.

 
 

  
  
[ora11g@rac1 DUMP]$ ls -lrt *.dmp|awk '{print $9}'|awk -F_EXT '{print $1}'|uniq  
  
  
TEST1  
  
  
TEST2  
  
  
TEST3  
  
  
TEST4 
 
 

Next, let's take a closer look at how many dump files correspond to each table.

You can use the following script B. sh to quickly obtain

 
 

  
  
ls -lrt *.dmp|awk '{print $9}'|awk -F_EXT '{print $1}'|uniq|xargs -i echo "echo {} \`ls -lrt {}_EXT_[0-9]*.dmp|wc -l \`>> tablst" >test.sh  
  
  
ksh test.sh  
  
  
rm test.sh  
  
  
[ora11g@rac1 DUMP]$ ksh b.sh  
  
  
[ora11g@rac1 DUMP]$ cat tablst  
  
  
TEST1 100  
  
  
TEST2 450  
  
  
TEST3 300  
  
  
TEST4 150 
 
 

In this way, the statistics of files are clear and clear.