Periodic OPS down on couchbase server 3.0.1 every hour

This issue is from https://forums.couchbase.com/t/periodic-ops-down-on-couchbase-server-3-0-1-every-hour/2398

@ingenthr,
I'm so sorry too late for filing on your issue tracker. I've attached files generated using cbcollect_info you mentioned. (collect.zip is too large to upload here, so I've separted two files, cbcollect_1.tgz and cbcollect_2.tgz)

Every 44, 48min, so many timeouts are occurred on cb clients .
I use cb java sdk v2.0.2 and I have 2 Clusters for joining data.

• There's no periodic throughput down like this on couchbase server 2.2
• And there's no batch job on all nodes.
• It occurs 44min and 48min every hour exactly no matter with gc and request loads (it acts like alarm ;()
• There's no network effects like high traffic periodically.
• I never set TTL on documents and never configure auto compaction option.