8.1.5.5. Memory Dashboard
The Memory Dashboard can be used to monitor the cluster memory usage and demand. Figure 54 shows a sample dashboard that has been created using four w
idgets. These widgets include: Figure 54. Cluster Memory Dashboard
The Generic Scoreboard widget has been configured with these metrics defined at the cluster level. The Resource cluster is used for the memory metric selections.
Table 3: Generic Scoreboard Configuration
Metric | Unit | Green Range | Yellow Range | Orange Range | Red Range |
Memory Granted | KB | 8000 | 8000–12000 | 12000–13000 | 13000 |
Usage | % | 50 | 50–75 | 75–80 | 80 |
Usage/Usable | % | 50 | 50–75 | 75–80 | 80 |
Contention | % | 50 | 50–75 | 75–80 | 80 |
Swap In Rate | KBps | 300 | 300–400 | 400–500 | 500 |
Swap Out Rate | KBps | 5 | 5–10 | 10–20 | 20 |
Note: These threshold/ranges are only examples. These values should be based on the cluster design threshold values and the customer requirements.
Figure 55. Cluster Memory Scoreboard Widget
The Heat Map is configured with three metrics, created from the cluster resource, as follows:
Swap Out Rate (KBps).
Swap In Rate (KBps).
Mem Swap Used (Kb).
The settings for the Heat Map widget are shown in the following table.
Metric | Value |
Attribute Kinds | Memory |
Tags to Filter | Cluster compute resource (select the correct cluster) |
Minimum Value | (Custom setting to environment) |
Maximum Value | (Custom setting to environment) |
Figure 56. Cluster Memory Heat Map Widget
The Metric Graph Widget can display line graphs of historical memory statistics. This example selects the following memory metrics to be displayed over a 24-hour period. Note that the historical time can be changed to up to one year and can also indicate the DT (dynamic threshold) values.
Table 18. Cluster Memory Metric Graph Settings
Metric |
Memory | Contention (%) |
Memory | Guest Active (KB) |
Memory | Granted (KB) |
Memory | Usage (%) |
Figure 57. Cluster Memory Metric Graph Widget