Release Notes - Couchbase Server - Version 2.0-beta-2 - HTML format

Bug

  • [MB-4135] - esc key causes Lost connection to server message
  • [MB-4392] - moxi returns two results when setting non-empty value for zero-length SET
  • [MB-4463] - Up sign icon is weirdly round in only one corner
  • [MB-4794] - Clients do not handle various views related errors
  • [MB-4865] - Value emitted for each key does not wrap around in the Dispaly
  • [MB-5759] - Excessive resource consumption with many connections to the streaming API (8091)
  • [MB-5772] - turn off compaction file validator in 2.0 rpm/deb/windows builds
  • [MB-5938] - UI needs to deal with very slow docloader (alternatively, docloader is too slow)
  • [MB-6209] - auto-failover should not fail over a node in the cluster if there are not enough number of replicas in the cluster
  • [MB-6460] - extra Couchbase Server Alerts are added on the page after each failures
  • [MB-6467] - Seeing uneven CPU utilization on cluster nodes, some nodes show high CPU (80-90%, expected) and some show very low (0.4 %)
  • [MB-6501] - source tarball make does not work because memcachetest make dist does not generate tar.gz file
  • [MB-6540] - Free disk space shows negative value
  • [MB-6563] - Replication ops/sec drops to 0, on stop/start load from source cluster. [Load with only creates looks okay, load with expired items--> replication rate drops very low.]
  • [MB-6586] - Replication rate may drop when the XDCR replication queue size becomes less than 500k items
  • [MB-6592] - [longevity] memcached hangs when aborting during swap rebalance operation and fails to restart ( exit 71 )
  • [MB-6612] - queries returns error {badmatch,{not_found,no_db_file}
  • [MB-6617] - [longevity] rebalance does not fail when 2 nodes down during rebalance
  • [MB-6619] - XDC: get latency is ~ 10 times higher on destination nodes
  • [MB-6625] - Delete/Recreate identical bucket on xdc destination cluster, results in replication on the remote cluster using previous replication id?(likely)
  • [MB-6626] - on EC2 fix the query link on the views screen to point to external IP address (instead of internal)
  • [MB-6627] - after Failed over 'node_id': ok button failover should be disabled, otherwise we can failover node many times
  • [MB-6628] - Seeing incorrect cache-miss-ratio ( very large ranging upto 500) during bidirectional XDCR.
  • [MB-6631] - Very small amount of items pending in XDC Queue on source cluster after replication has completed.
  • [MB-6633] - querying for views during rebalancing through node's REST api before it has any active vbucket return ""error":"no_active_vbuckets"
  • [MB-6637] - make replication batch size pluggable
  • [MB-6638] - beam.smp ( erl_bestfit_alloc.c:435 ) segmentation fault when rebalancing a new node to a cluster with 2 buckets and 1 design doc per bucket
  • [MB-6640] - Indexing progress stuck at 100%
  • [MB-6643] - Very slow replication rate( less than 100 on average) with 2 unidirectional replications between 2 clusters.
  • [MB-6645] - Unit test failure in expiry_no_items_warmup
  • [MB-6647] - Sporadic failure in "duplicate items on disk" unit test
  • [MB-6649] - beam.smp memory usage grows to 2 GB when xdcr feature is enabled and rebalancing is in progress
  • [MB-6650] - [windows] Queries with stale=false shows errors {case_clause,{error,eacces}} ( rebalance test of 4->3 nodes when queries are run during rebalancing is hung )
  • [MB-6653] - [windows] Unable to update ddoc(1 bucket, single node)
  • [MB-6661] - We need to return a cas for in the delete packet
  • [MB-6663] - Cli command "./cbclusterstats -h" returns error
  • [MB-6666] - User-triggered view compaction fails on few nodes with error {"error":"unknown_error","reason":"view_undefined"}
  • [MB-6673] - mac 2.0 builds ask from user whether they want to install command line tools under /usr/bin
  • [MB-6674] - beer sample data spatial points does not work
  • [MB-6675] - modify brewery_beers view design doc
  • [MB-6679] - web UI timeout leads to loss of user work/data
  • [MB-6681] - [windows] items are not drained to disk after rebalance 6->7 nodes
  • [MB-6683] - [system test] beam.smp ( in memcpy () from /lib64/libc.so.6 ) segmentation fault when start couchbase server
  • [MB-6689] - [consistent views enabled] rebalance 2->4 failed with reason {{function_clause, [{couch_set_view_util, missing_changes_count
  • [MB-6695] - null values in replication_commit_time stats
  • [MB-6700] - Very Uneven distribution of replicated data on a 2 unidirectional xdcr setup during initial loading phase.
  • [MB-6705] - Slight change to the Documentation page, possibly?
  • [MB-6708] - Docs not getting persisted to disk with consistent views enabled
  • [MB-6710] - Deleting bucket on a cluster gives error "Some nodes are still deleting bucket"
  • [MB-6711] - size of one vBucket is 10 GB - Rebalance exited with reason replicator_died( many retries to notify CouchDB of update : vbid=294 rev=1)
  • [MB-6713] - enable consistent views by default for 2.0 beta refresh builds ( and remove the internal setting dialog )
  • [MB-6715] - Flushing the same bucket twice gives error - "Failed to flush bucket"
  • [MB-6723] - Rebalance exited with reason"missing_checkpoint_stat" during empty bucket rebalancing on source and destination cluster.
  • [MB-6724] - XDCR : Data is unevenly replicated on destination cluster. 2 of 5 nodes, not getting any replicated data.
  • [MB-6725] - When Memcached Client (python) delete an item, the item's cas returned by delete operation is NOT "0" (since build 1758)
  • [MB-6728] - [system test] view page shows spinning circle when no bucket created
  • [MB-6731] - cbdocloader is broken since build 1762
  • [MB-6732] - XDCR: Replication not as expected on build 1762
  • [MB-6736] - Querying a view on a single node gives inconsistent_state error if bucket is empty
  • [MB-6748] - XDCR: Items not being replicated to non-master nodes on the destination
  • [MB-6749] - Very slow XDC changes queue and incorrect "docs_replicated" number on Source cluster on rebalances on source/destination clusters.
  • [MB-6751] - Incorrect error message from 'vkey' when key does not exist
  • [MB-6757] - cbdocloader can't load views for a sasl bucket created by user build 1776
  • [MB-6759] - Querying a view on a 4 node cluster gives error "Badarg error in HTTP request"
  • [MB-6761] - XDCR Error Logging for Flush bucket and Delete bucket on source/destination
  • [MB-6764] - cbepctl --help doesn't show options on how to set expiry_pager
  • [MB-6767] - Unable to load all 5 Outgoing XDCR UI stats, when seen by server
  • [MB-6769] - Very slow rebalance progress while rebalancing-In on Source cluster during Unidirectional XDCR.
  • [MB-6772] - From the settings tab: user is able to load only one sample data at a time using the sample loader instead of loading both at the same time
  • [MB-6774] - When Quitting Couchbase Server on Mac, epmd is still running
  • [MB-6777] - Query request for _design/_info returns error no_db_file while view compaction is in progress(consistent views enabled)
  • [MB-6779] - Apparently ep-engine continues to try to write to old revision of vbucket db continuously breaking persistence and rebalance (was: [consistent views enabled] Rebalance hungs, Write commit failure appears)
  • [MB-6781] - creating and querying one spatial index on mac returns error,system_limit because mac file descriptor limit is not reset to 10k
  • [MB-6783] - increase number of file descriptors allowed for couchbase server ( couchbase user ) for Mac installation
  • [MB-6784] - Sample databases beer-sample/gamesim-sample fail to load any data
  • [MB-6786] - high cpu usage reported after couchbase server 2.0 installation ( errors when before_header,after_header,on_file_open)
  • [MB-6787] - XDCR: source nodes crashed during initial load
  • [MB-6792] - [consistent views]Rebalance 4->1 failed with reason view_undefined
  • [MB-6796] - perf - 10x slower rebalance performance with consistent view turned on
  • [MB-6805] - [windows] rebalance test of 4->3 nodes when queries are run during rebalancing is hung
  • [MB-6806] - memcached crashed in flush_buffer -> Rebalance exited with reason {mover_failed,{badmatch,{error,closed}}}
  • [MB-6810] - moxi and memcached exited during rebalance
  • [MB-6813] - XDCR Replication fails with views enabled, with small front end load.
  • [MB-6814] - Rebalance exited with reason {mover_failed,{badmatch,{error,closed}}} during rebalances on empty buckets on both clusters with XDCR.
  • [MB-6815] - [system test] view compaction hang at 98 percents
  • [MB-6816] - Rebalance on target cluster of XDCR is still causing source remote_clusters_info to crash sometimes causing source cluster rebalance to crash
  • [MB-6817] - Views page does not load when no buckets exist
  • [MB-6819] - Rebalance exited with reason {not_all_nodes_are_ready_..}, after upgrade 1.8.0->2.0 when add 2.0 node to mix cluster
  • [MB-6828] - Views/Spatial views stay "loading" when a bucket is deleted
  • [MB-6840] - GAT not populating proper error information
  • [MB-6844] - [system test] failed to delete data, cannot locate database file
  • [MB-6847] - Data loss during upgradation from 1.8.1 to 2.0.0
  • [MB-6904] - System testing for XDCR: Entire cluster is gradually down/pending, seeing few cores on beam.smp - Sig Abort
  • [MB-6942] - Spatial views return debug information and wrong number of total_rows
  • [MB-6974] - erlang FULL_SWEEP setting needs override to 512 instead of erlang's default
  • [MB-6985] - Regression: Delete with CAS is broken

Improvement

  • [MB-6753] - Provide the timestamp of TAP backfill start
  • [MB-6776] - RESTful flush should be disabled by default
  • [MB-6804] - Ignore effectless indexable/unindexable state transitions
  • [MB-6866] - Use different kv and kp node thresholds for views
  • [MB-6868] - Don't use couch_set_view:get_group_info/2
  • [MB-6988] - Add more data to queries with ?debug=true, and make it work equally for reduce queries

Story

  • [MB-6169] - upgrade - use larger batch sizes during 1.8 to 2.0 offline upgrade for performance

Edit/Copy Release Notes

The text area below allows the project release notes to be edited and copied to another document.