Slight replication regression in one of the XDCR tests (5 -> 5), 1G, DGM test

Build : 6.6.5-10068

Test :  Avg. initial XDCR rate (items/sec), 5 -> 5 (2 source nozzles, 4 target nozzles), 1 bucket x 1G x 1KB, DGM

http://showfast.sc.couchbase.com/#/timeline/Linux/xdcr/init_multi/all

html report : http://cbmonitor.sc.couchbase.com/reports/html/?snapshot=titan_c1_665-10068_init_xdcr_908d&snapshot=titan_c1_663-9808_init_xdcr_fdcd

All other tests look good.

Looks like 6.6.5 may be using a bit more memory.