GRID - Cluster Gridhosts/Gridbhosts Stats

This graph shows the amount of time that the RTM Grid API is taking to collect host and batch host statistics from each LSF cluster. Lower values are better. In this example response times for gridhosts and gridbhosts are generally between 100 to 400 milliseconds. Gridhosts is using the LSF API to extract basically the same information as is provided by the lshosts command and storing data to the database about hosts. Gridbhosts is gathering batch-related statistics and storing them to the database similar to what the bhosts command would show.

The following graph shows that the response times are consistently less than 1 second. However, there was a period around week 15 where the response was slower. This result can be the result of either heavy workload, or some other issue with the system.

Figure 1. GRID - Cluster Gridhosts/Gridbhosts Stats