investigate memory usage during diag collection (was: goxdcr, ns_server crash while grabbing cbcollect, restart and crash again)

Build
------
4.0.0-1697

goxdcr crashed on .171 while collecting cbcollect-info.Subsequently ns_server crashed and then came up.

Port server ns_server on node 'babysitter_of_ns_1@127.0.0.1' exited with status 137. Restarting. Messages: (repeated 4 times) ns_log000 ns_1@172.23.107.170 17:19:33 - Mon Mar 30, 2015
Port server goxdcr on node 'babysitter_of_ns_1@127.0.0.1' exited with status 1. Restarting. Messages: github.com/couchbase/goxdcr/gen_server.(*GenServer).run(0xc2084ce280)
/home/couchbase/jenkins/workspace/sherlock-unix/goproj/src/github.com/couchbase/goxdcr/gen_server/gen_server.go:78 +0x839
created by github.com/couchbase/goxdcr/gen_server.(*GenServer).Start_server
/home/couchbase/jenkins/workspace/sherlock-unix/goproj/src/github.com/couchbase/goxdcr/gen_server/gen_server.go:68 +0x70
[goport] 2015/03/30 17:19:01 /opt/couchbase/bin/goxdcr terminated: exit status 2

After two mins, ns_server crashed again on the same node.

Pls see screenshot attached of the UI log.