Fyi, Nano node bug currently causing significantly increased confirmation times
We have recently observed higher than expected node resource usage on the Nano network. This includes more CPU, disk IO and bandwidth being consumed by elections than is typical compared to the number of transactions being confirmed.
We are currently investigating the root cause and working to resolve these conditions. Areas of investigation include frontier confirmation, sequential voting and elections, related backtracking and how voting and confirm_req are issued for active elections.
For those looking to conduct broad tests on the main network it is recommended those be done on the beta network or temporarily postponed, since the results will be inaccurate.