Percona's pt-table-sync: how to run on more than one table?

Question

In the command line, this will successfully update table1:

pt-table-sync --execute h=host1,D=db1,t=table1 h=host2,D=db2

However if I want to update more than one table, I'm not sure how to write it. This only updates table1 as well and ignores the other tables:

pt-table-sync --execute h=host1,D=db1,t=table1,table2,table3 h=host2,D=db2

And this gives me an error:

pt-table-sync --execute h=host1,D=db1 --tables table1,table2,table3 h=host2,D=db2

Anyone have an example of how to list the '-tables'... so that it successfully update all the tables in the list?

Answer 1

The --tables option seems to be incompatible with the DSN notation, you get this error:

You specified a database but not a table in h=localhost,D=test.
Are you trying to sync only tables in the 'test' database?
If so, use '--databases test' instead.

As suggested in that error message, you can use --databases and then you can use --tables successfully.

For example, I created tables test.foo and test.bar, filled each with three rows, then deleted the rows from test.bar on the second server dewey.

I ran this:

$ pt-table-sync h=huey h=dewey --databases test --tables foo,bar --execute --verbose

# Syncing h=dewey
# DELETE REPLACE INSERT UPDATE ALGORITHM START    END      EXIT DATABASE.TABLE
#      0       0      3      0 Chunk     15:26:15 15:26:15 2    test.bar
#      0       0      0      0 Chunk     15:26:15 15:26:15 0    test.foo

It successfully re-inserted the 3 missing rows in test.bar.

Other tables in my test database were ignored.

Answer 2

This is an old question, but I searched everywhere for an answer. pt-table-sync only does one table. There is no tool that does the same thing to a list of tables or a full database schema. Specifically I want to run a Live server and be able to sync back to a Staging server, then edit code and files in the Staging server without fear of messing up Live or being overwritten by Live... and I want it to be free :)

I ended up writing a shell script called mysql_sync_live_to_stage.sh as follows:

#!/bin/bash
# sync db live to staging

error_log_file='./mysql_sync_errors.log'
echo $(date +"%Y %m %d %H:%M") > $error_log_file

function sync_table()
{
    pt-table-sync --no-foreign-key-checks --execute 
        h=DB_1_HOST,u=DB_1_USER,p=DB_1_PASSWORD,D=$1,t=$3
        h=DB_2_HOST,u=DB_2_USER,p=DB_2_PASSWORD,D=$2,t=$3 >> $error_log_file
}

# SYNC ALL TABLES IN name_of_live_database
mysql -h "DB_1_HOST" -u "DB_1_USER" -pDB_1_PASSWORD -D "DB_1_DBNAME" -e "SHOW TABLES" | 
        egrep -i '[0-9a-z\-\_]+' | egrep -i -v 'Tables_in' | while read -r table ; do
    echo "Processing $table"
    sync_table "name_of_live_database" "name_of_staging_database" $table
done

# FIX Config Settings For Staging
echo "Cleanup Queries..."
mysql -h "DB_2_HOST" -u "DB_2_USER" -pDB_2_PASSWORD -D "DB_2_DBNAME" 
    -e "UPDATE name_of_staging_database.nameofmyconfigtable SET value='bar' 
    WHERE config_id='foo'"
mysql -h "DB_2_HOST" -u "DB_2_USER" -pDB_2_PASSWORD -D "DB_2_DBNAME" 
    -e "UPDATE name_of_staging_database.nameofmyconfigtable SET value='bar2' 
    WHERE config_id='foo2'"
echo "Done"

This reads a list of table names from the live site then executes a sync on each one via the do loop. It goes through the list alphabetically, so I recommend keeping the --no-foreign-key-checks flag.

Its not perfect... It won't sync tables that don't exist in both databases, but when combined with a "git pull -f origin master" I get a complete sync in a couple minutes.