Provide "rampup" rule/option
maxfischer2781 opened this issue · comments
Max Kühn commented
When restarting a server while the cluster is under load, the restarted server naturally starts with low load. This causes an initial low perf
report and likely leads to the server getting most if not all new requests.
It would be useful to allow a "rampup" functionality that starts reporting a high load and only approaches the real load over time. This could be a grammar rule (e.g. rampup(pcpu, 5m)
) or general option (e.g. --rampup 5m
).