solidDB Help : solidDB Grid : Working with grids : Monitoring a grid
  
Monitoring a grid
Note Check the solidDB Release Notes for any limitations that are associated with using a grid in the current release.
You can use the diagnostics and monitoring features that are available for individual nodes to monitor grid nodes, see Monitoring solidDB.
However, monitoring a grid is more complicated for the following reasons:
▪ a grid consists of multiple solidDB instances,
▪ individual nodes can legitimately join or leave the grid during monitoring.
The first task is to identify the nodes that are, or were, part of the grid. Then, regular monitoring commands must be executed for each grid node and the results collected for consolidation.
The following points should be noted when monitoring a grid:
▪ For some admin commands, you can execute the command in all current grid nodes by using the 'global' keyword, see ADMIN COMMAND syntax.
▪ To see a list of grid-specific administration commands, run the following command:
ADMIN COMMAND 'grid commands'
For more information, see GRID
▪ The nodes that are currently part of the grid can be found out from the system table SYS_GRID_NODES, see SYS_GRID_NODES.
▪ Diagnostic information about former nodes or nodes that have been down for a long period of time might be relevant for grid monitoring.
▪ If a grid is used in a container, diagnostic data stored in output files in the container is lost if the container fails.
Go up to
Working with grids