I have some large MySQL production tables that I need to dump so that I can run some analysis in Hadoop. What best practice should I know about dumping and querying production databases? Do I need to worry about affecting production performance if I just dump / (no one else writes)?
Yes, you need to worry about it being locked while reading MyISAM tables No modifications will be allowed, just another reads. InnoDB does not have this problem, but if you are checking for large sets of data it will be locked, the dump will lock any type of table in almost all cases.
What is usually done is that you set up a slave database and do all your inquiries on it, even for a purpose reporting just set up a slave you set the master / slave configuration Will need to stop the MySQL server. But once you do it once, you can use the slave to make other slaves.
Comments
Post a Comment