When viewing some metrics in OpsCenter, the latency figures appear to be higher than normal
When looking at latency metrics such as read / write request latency for nodes the figures appear to be much higher than normal when running tests with a benchmarking tool such as cassandra stress. When using nodetool to check for latency such as "cfhistograms" and "proxyhistograms" the same figures are not seen.
The problem was caused by a known issue in OPSC-3905 where the metric graphs were reporting the wrong values
Upgrade to OpsCenter 5.0.2 or later