Well, maybe cat table is a good one to try with (but I’m not very happy in rebooting mysql each time the admin change something in cat table)

With regards to MySQL not doing anything to aid the issue, you need to remember that InnoDB is the product of Innobase, an Oracle company, and it’s up to them how the datafiles are managed. Try Falcon or Maria

Like you, I don’t see any benefit in the second backup. However, I’d do it differently depending on space and resources available.

Stop connections to the database.
Then either…
Dump the database data only, one datafile for each table.
Dump the database structure only with “drop tables” enabled.
Reload the database from the structure file, this will drop and re-create the tables.
Load the data from the dump files.

OR…
for each table, create a new table, table_new.
The insert into table_new select * from table
rename table to table_old,table_new to table.

The benefit of the second approach is that if you’re clever, you can tailor the inserts to include only data that your system needs initially, to speed deployment, then load across older data as time goes on, and that it can be done (at a push) on a live system, but it is, ultimately, slower.

FWIW, I’m just hitting this issue after several years of use but it is a very serious shortcoming. Posts over 5 years old on the MySql web site but nothing yet. Why?

Step 1: Installing mytop pre-requisites

6. Once Perl DBI is installed, you need to install Perl DBD::mysql the mysql driver that goes with the DBI module. This is is particularly nasty to install. Check if you already have it by running: rpm -qa | grep DBD; if you got it you are in good shape.

Here you are in trouble. You are “filtering” the view using a column. This forces MySQL to create a temporary table every time you execute this query! Well, not exactly every time if you call the same query again again (Delete repeated word). But most of the time, when you execute this query, it creates a temporary table and thus it costs a lot of execution time.

Problem Statement:
When you are developing Web applications that uses a MySQL database, sometimes you need to monitor MySQL processes (threads) to see if any query is taking unusally (unusually) long time or not.

I have gone through this. Please check the following lines if they need a revision.

“This is a very desirable solution as the burden of database size issue falls on the Web application and not the database itself.”- I thought it would be “..and not on the database itself.”

“When you recreate a table to exists as a union of two or more merged tables”- I thought it would be ‘exist’

“so that INSERT traffic completes less with your SELECT, UPDATE, DELETE traffic for older records.” - Would it be “competes”?

“Since your Web application only deal with the order table itself..”- Would it be ‘deals’?

Thanks.

I have gone through this and found the following lines to be confusing. Please revise.

“You need an automated of loading the memory tables;”

“Now create a text file called /etc/my.memory.tables.init.sql file to to have a statement”

“this will make sure the statements load data from the correct database.”

Thanks.

I have gone through the article and found this ‘bottlenack’.
Also I felt, the following phrase needs a revision.
“which provides different data when run”

Thanks.