Problem : I have one master and various slave on EC2, all running with the innodb_file_per_table as OFF.
We have done some testing work on one of the slave say server A.and it takes a huge amount of disk space due to out test work on that slave. (Approx 300GB increased due to our test work) as dropping the database would not help us due to innodb_file_per_table OFF.
ibdata is on LVM. datadir is on /EBSroot innodb logs are on some other location
The total ibdata size is approx 600GB due to this test work it increased to 900GB.
Now i want to reclaim the space.I am thinking as
delete the ibdata from server A.and take the snapshot backup from the one of the slave and restore on this slave A.
but i am not getting the correct steps.For How to take the backup from one of the slave and restore on this slave A.
i have studied alot on how to take snapshot backup of MySQL i found some Caveats as : Handling data spread on multiple volumes (DB logs on separate LV or DBs spread across multiple LVs).
So how can do this without any issues.because my ibdata/logfiles/datadir are all on the different locations.I have to also restart the replication as well.
I need the best possible way with minimum downtime.