uber / cherami-server

69%

master: 67%

LAST BUILD ON BRANCH release-v1.8.0-rc3

branch: release-v1.8.0-rc3

CHANGE BRANCH

x

Reset

• release-v1.8.0-rc3
• DLQConsumedMerge
• DLQConsumedMergeHotPatch
• Update_cherami-client_to_v2.4.0
• ackmgr-stop-ticker
• active_zone_switch
• active_zone_update
• admin
• admin-command-store-list-extents
• alert
• api-migration
• auth
• autoLocker
• benchmark
• bug
• bugs
• bytes_counter
• cache
• cassandra-auth
• cassandra-read-consistency-increase
• cdb
• cg-cleanup
• cg-dest-uuid
• check-dlq-zero-uuid
• cherami_EOL
• cleanup-dlq-dest
• cli-fix
• cli-showcg
• cli_fix
• cli_helper
• client_side_drain
• cmq
• cmq-delete-cql-consistency-all
• cmq-gc
• cmq-num-delay
• cmq-repair
• config
• config-cli
• configExtent
• configmgr
• consist
• consistency
• consistency-level-configurable
• consumer-group-deletion
• controller-log-metadata-errors
• controller_bug
• controller_drain
• correct-startfrom-units
• coverage
• createdst_urn
• cut_release
• deadlk_ext
• delay-and-skipolder
• delay-metadata
• demo-deploy-scripts
• dev
• disable-TestQueueDepth
• disable-TestStartFromWithCassandra
• disable-flaky-timer-tests
• disable_read_perm
• distance
• dlq-cleanup
• dlq-extent
• dlq-zero-uuid
• docker
• docker_fix
• drain
• enable-timer-test-logging
• extent-down-metadata-read-failure
• failovermgr_metrics
• fast_drain
• fix
• fix-TestGetOutputHostsKafka-panic
• fix-cfg
• fix-cli
• fix-cli-defaults
• fix-cli-help
• fix-cnsmid-tag
• fix-commonkafkaconfig-race
• fix-delete-destination-kafka-cluster
• fix-elk
• fix-enqueue-timestamp-type
• fix-fmt
• fix-frontend-startfrom
• fix-getoutputhosts
• fix-logs
• fix-offsets-bug
• fix-queue-depth-replica-stats
• fix-race-storehost-test
• fix-replicate-extent-test
• fix-retmgr-empty-extents
• fix-retmgr-empty-sealed-extents
• fix-retmgr-logs
• fix-storehost-elk-logging
• fix-test-panic
• fixQueueDepthInfo
• fix_FD
• fix_cassandra_config
• fix_controller
• fix_drain
• fix_integration_test
• fix_lint
• fix_lock
• fix_race
• fix_rocksdb_embed
• fix_shutdown
• fix_tchannal
• fix_test
• fixes
• flack_test
• fmt
• g
• get-latest
• getinputhosts
• getoutputhosts
• glide
• glideUp
• go-1.8
• go1.9
• gocql-tweaks
• govet
• govet_audit
• graceful_drain_input
• inputDConfig
• inputEmpty
• inputEmptyMinimal
• input_admin
• input_misc_perf
• input_size_seal
• input_upgrade
• integtest
• kafka-liveness-ut2
• kafka2
• kafkaAcks
• kafkaCLI
• kafkaClustersConfig
• kafkaRead
• kafkaSchema
• kafkaStream
• kfc-extentmon
• kfc-extents-increase
• kfc-getoutputhosts
• kfc-integration-test
• kfcIntegrate
• kfcOutMetric
• latency
• lint
• lintay
• listen
• lockmgr-fix
• log
• log-metadata-read-failure
• logs
• master
• md-mgr
• metadata-client-logs
• metadata-client-module-tag
• metadata_listrequest_ValidateAgainstPathTable
• metametrics
• metametrics-log
• metric
• metrics-input-bytes
• metrics-output-bytes
• metricsTesting
• mingmin/expose-GetDefaultKafkaMessageConverter
• mingmin/expose-kafkaAddresser
• mingmin/fix-lint
• minor_tuning
• minus_kafka
• misc-fix
• misc-scripts
• mock
• more-log-fix
• msgcache-timer-fix
• multi_zone_cg
• multi_zone_cg_extent
• multi_zone_dest_check
• multi_zong_dlq
• mz_fix
• mz_upgrade
• nodestatus
• output
• output-fix
• output-fix-tail-loss
• outputStuck
• output_admin
• output_leak_fixes
• output_misc_perf
• output_msg_cache_cfg
• outputhost-delay-and-skipolder
• parallelize_tests
• phantom-extent
• port
• publatency-threshold
• purge-latency-log
• queueDepth
• queueDepthInfo
• queue_depth_tabulation
• r_b
• r_bug
• r_f
• r_r
• r_race
• read-level-resume
• real_integration_fix
• really_fix_FD
• receiveMsgBatch
• reduce_sizes
• reenable-replicate-extent-test
• refactor
• refactor-replica-stats-reporting
• release-v1.18.0-rc3
• release-v1.23.0-rc0
• release20170626
• reminder
• remove-printf
• remove_s3
• remove_unused
• repl-conn-leak
• replica-rollover-metric
• replication_alerts
• replicator_busy_loop
• replicator_counter
• replicator_fd_leak
• replicator_fix
• replicator_fix_1
• replicator_fix_deadlock
• replicator_host_updater
• replicator_hotfix
• replicator_monitoring
• replicator_reconcilation_fix
• replicator_reconcile_fix_1
• replicator_reconcile_issue
• report-atleast-once
• retention-kafka
• retmgr-non-primary-controller
• retry
• revert
• revert-maxprocessingtime-change
• revert_a
• revert_processing_time
• rocksdb
• rocksdb-5.0.2-trusty
• rp
• rp_fix
• rr_f
• schema
• seal-check-dlq
• seal-extents
• seal-seq
• security10
• security11
• security12
• security15
• security17
• security4
• security7
• security8
• security9
• selectNextExtentsToConsume
• service_start_wait
• slow_down
• smartRetryAutoDisable
• smart_retry
• spaceMon
• srLRU
• startfrom-test-fix
• store-kafka-dlq
• store_counter
• store_upgrade
• storehost-logs
• storehost-notify-optimize
• storehost-readonly
• storehost-readonly-notify
• storehost-readonly-space-threshold
• storehost-replication
• storehost-started
• storestatus
• storetest
• syncset
• tags
• test-kafka-travis
• test-v1.26.0
• test-verbose-logging
• testFix
• testbuild
• throttle
• throttler
• tightenAckAggregate
• timeout
• try_FD_leak_again
• unload-all-cg
• untagged
• untagged-f5472780099c97841c0d
• update
• update_cg
• update_checkauth
• update_client
• update_rocksdb
• update_thrift
• upgrade_support
• upstream_patches
• use_prebuilt_rocksdb
• v0.1.0
• v0.2.0
• v0.2.0-rc1
• v1.16.0-output-hotfix
• v1.16.0-rc1
• v1.16.1-rc1
• v1.17.0-rc1
• v1.18.0-rc1
• v1.18.0-rc3-patch
• v1.19.0-rc1
• v1.20.0-rc0
• v1.20.0-rc1-hotfix0
• v1.21.0-rc2
• v1.22.0-rc1
• v1.22.0-rc2
• v1.22.0-rc2-hotfix1
• v1.23.0-rc0
• v1.23.0-rc0-replicator_hotfix
• v1.24.0-rc0
• v1.25.0-rc0
• v1.25.0-rc3-hotfix1
• v1.26.0
• v1.26.0-rc0
• v1.26.0-rc1
• v1.27.0-rc0
• v1.27.0-rc1
• v1.27.1
• v1.28.0
• v1.28.1
• v1.8.0-rc3
• validate_input
• write_to_cache

Build # 1252

Build Type

push

travis-ci

Committed by KiranRG

Commit Message

Fix drain (#125)

* Make sure the drain is sync

We were not actually waiting for the drain to finish before returning
but rather we just stopped the write pump (inflight messages are still
flowing).
This will mean the controller will simply seal the extent as soon
as we return.

* Create a separate bucket for drain and startfrom tests

* Make sure the drain timeout is plumbed through

* Fix integration test

Make sure to check if we have drained already even if we don't
get any inflight messages.

* Revert the unintentional config change

* Make cassandra success test to be part of ParallelE suite

Run Details

19934 of 28864 relevant lines covered (69.06%)

3321.03 hits per line