Cluster Metrics

The Cluster Metrics page shows consolidated statistics for all segment hosts in the Greenplum cluster. Master and standby master hosts are excluded from the metrics.

The charts display metrics for the last time period set by the control in the top right corner of the screen.

Use the Show/hide Charts control to choose which metrics to display.

Hover over any of the charts to see values for the metrics at a point in time in pop-up boxes. The charts are synchronized so that hovering over any chart shows the same point in time in all charts.

The current value of a metric is shown in the upper right corner of its chart.

On charts with multiple metrics, toggle the display for a line on or off by clicking the line’s label in the legend at the top right of the chart. At least one line must be displayed. All lines are redisplayed at the next quantum interval.

The page has charts for the following metrics:

The number of queries running and the number of queries queued to run.
The percentage CPU used by system processes and the percentage CPU used by user processes.
Percentage of memory in use.

Memory is calculated as follows:

Total = MemTotal
Free = MemFree + Buffers + Cached
Used = MemTotal - Free

Disk I/O
Disk read and write rates in megabytes per second.
Network I/O read and write rates in megabytes per second. Network metrics include traffic over all NICs (network interface cards), including internal interconnect and administrative traffic.
System load average for 1-minute, 5-minute, and 15-minute periods.
Percentage of swap space used.