Couchbase Server

Performance Status

06/25/2014

Bug Snapshot

Last Wednesday:
  • Test blockers: 1
  • Blockers: 6
  • Critical: 5

Today:
  • Blockers: 3
  • Critical: 8

DefeRred / MOVED


  • MB-11143: Avg. BgFetcher wait time is 3-4 times higher on a single HDD in 3.0
  • MB-10679: It takes almost 10 minutes to stop Couchbase service in setup with 10 empty buckets
  • MB-9676:  Dramatically increasing latency of SET operations during rebalance tests

XDCR

  • MB-11435: 1500-2000% CPU utilization by beam.smp on source nodes in XDCR scenarios (was ~500% in 2.5.x)
  • MB-11434:  600-800% CPU utilization by memcached on source nodes in XDCR scenarios (was <100% in 2.5.x)
  • MB-11412: memcached memory rapidly grows to OOM during initial XDCR
  • MB-11382:  XDCR: default replicators per bucket is now 16 (earlier: 32)
  • MB-10437:  XDCR replication rate drops almost to zero in presence of light write workload on src side

Views

  • MB-11486: Erlang memory usage increases to 50GB in 10 minutes after "unexpected_binary", node obviously fails afterwards
  • MB-11464: Initial indexing of 20M docs gets stuck
  • MB-11461: 2x regression in latency of stale=update_after queries
  • MB-11387: Latency of stale=false queries easily reaches 1 minute and more (moderate workloads)
  • MB-11384:  Erlang memory usage in cases with views is 4x higher than in CBEE 2.5.1
  • MB-10273:   View compaction doesn't catch up in basic non-DGM tests with view queries

KV

  • MB-11474: 50% regression in KV rebalance
  • MB-11405: ~2400% CPU consumption by memcached during ongoing workload with 5 buckets
  • MB-11363:  Rebalance after failover (delta recovery mode) fails at the early beginning
  • MB-11347: mem_used stat differs from memory usage reported by tcmalloc and OS
  • MB-11005:    Erlang memory in KV test case with 10 buckets is too high (up to 30GB, leading to OOM situation)
  • MB-10771:  Delta recovery is slower than full recovery. Meanwhile several performance issues encountered.

    • Windows

  • MB-9825: Rebalance exited with reason bad_replicas

    • n1ql

    • MB-11141: SELECT COUNT(*) ... WHERE should avoid memcached operations
    • MB-11140: SELECT DISTINCT scans all docs
    • MB-11048: Range queries result in thousands of GET operations/sec

    BACKLOG


    • Knowledge transfer

    CBIT ZONE


    • CBIT-1182: client machines for regression tests

    Performance Status

    By Pavel Paulau

    Made with Slides.com

    Performance Status

    • 241

    More from Pavel Paulau