When importing data to a Clustrix system it is best practices to use the clustrix_import tool. This is because clustrx_import has been designed to take advantage of the parallel nature of clustered databases to import MySQL dumps into a Clustrix Cluster. Importing mysqldumps with the mysql client results in a single-threaded insert, which will be much slower than clustrix_import.
Detailed information on importing data to Clustrix can be found on our documentation site: Importing Data
If your import is running slowly or failing there are a few things to look at:
- You may get much faster import speed if you are running clustrix_import locally from a node, rather than from outside the cluster.
- Even though a password for the root user is not required when accessing sql from localhost, it is required for clustrix_import.
- We highly recommend running clustrix_import in a screen session so that if you are disconnected from the node, your import will continue to run.
- If you have been using triggers in MySQL you will want to run the mysqldump with the "--skip-triggers" argument.
If you are still having problems, please let us know, and include the clustrix_import output to help us understand where/how things are slow.