Monitoring MySQL Cluster Database Server

eG Enterprise provides an exclusive MySQL Cluster monitoring model that runs quick health checks on the MySQL database server at configured intervals, and proactively alerts administrators to potential bottlenecks to the performance of the server.

Figure 1 : The layer model of the MySQL Cluster database server

Using the model depicted by Figure 1, administrators can determine the following:

  • Is the MySQL Cluster database server running? If so, how many primary and secondary nodes are connected to the cluster?
  • Are the nodes in the target MySQL Cluster database server available? If so, how quickly does each node respond to user requests? How many nodes are unresponsive/unreachable?
  • Is the node overloaded?
  • Are clients able to connect to the node in the cluster , or are there too many connection failures?
  • Are connections been closed properly? Are there an unusual number of open connections to the node?
  • Is the query cache been utilized optimally?
  • Has adequate memory been allotted to the cache?
  • Is the key buffer cache utilized well?
  • Is query execution efficient, or do queries need to be optimized for better performance?
  • Are rollbacks kept at a minimum?
  • Should the sort_buffer be increased?
  • How is the overall locking activity on the server? Are too many requests waiting to acquire locks?
  • Are there too many long running queries? Which queries are running for a time duration longer than the configured time limit?

This section will deal with the first four layers of the layer model only, as the rest of the layers have been discussed elaborately in the Monitoring Unix and Windows Servers document.