Add sequential test method execution and support for HZ_LICENSEKEY env variable [CTT-1276]

[k-jamroz]

Adds sequential, round-robin test operations execution (see readme) and support for HZ_LICENSEKEY environment variable in perftest run

In some cases difference between sequential and random execution can be meaningful. For example IMap.delete is cheaper when the entry does not exist. This was demonstrated in the below example benchmark using deleteNewCurrent and setCurrent methods. The only difference between runs was sequential setting. The absolute numbers do not matter (the environment was quite weak 2 members) but the difference in latency (especially lower percentiles) and throughput is clearly visible.

run_label	random	random	sequential	sequential
benchmark	mapnative.deleteNewCurrent	mapnative.setCurrent	mapnative.deleteNewCurrent	mapnative.setCurrent
10%(us)	304.64	520.70	509.95	517.63
20%(us)	326.65	538.62	527.36	535.55
50%(us)	499.20	581.63	568.83	578.05
75%(us)	577.53	632.32	613.38	624.13
90%(us)	646.14	696.32	667.13	679.42
95%(us)	696.32	741.89	710.14	722.94
99%(us)	873.98	949.25	880.13	890.88
99.9%(us)	2,537.47	2,611.20	2,584.57	2,537.47
99.99%(us)	5,718.02	6,660.10	7,139.33	7,045.12
max(us)	19,120.13	19,677.18	21,118.97	21,086.21
operations	4,385,115	4,384,139	4,011,919	4,011,916
duration(ms)	117010	117010	117010	117010
throughput	37,476.41	37,468.07	34,286.98	34,286.95