Details
Description
A number of times I've seen people try to interpret sudden jump in server quota, drop in per-node disk usage, increase in hit ratio or other statistic change which is caused by the number of nodes in the cluster changing. Without the knowledge of exactly when the number of nodes changed (or indeed that a change occurred at all) it can be very hard to make sense of such changes.
To improve this, we could add a new stat to the GUI - #nodes - a count of how many nodes are in the cluster. While this is of limited value on the "live" view, it becomes very useful when viewing the day/week/month view as users can correlate other changes to the system against when the node count was modified.