8b52787114
Stream load with no data will abort txn ( #735 )
...
1. stream load executor will abort txn when no correct data in task
2. change txn label to DebugUtil.print(UUID) which is same as task id printed by be
3. change print uuid to hi-lo
2019-04-28 10:33:50 +08:00
062f827b60
Add attachment in rollback txn ( #725 )
...
1. init cmt offset in stream load context
2. init default max error num = 5000 rows / per 10000 rows
3. add log builder for routine load job and task
4. clone plan fragment param for every task
5. be does not throw too many filter rows while the init max error ratio is 1
2019-04-28 10:33:50 +08:00
192c8c5820
Fix bug that data consumer should be removed from pool when being got ( #723 )
2019-04-28 10:33:50 +08:00
8474061d63
Add some logs ( #711 )
2019-04-28 10:33:50 +08:00
fbbe0d19ba
Change the relationship between txn and task ( #703 )
...
1. Check if properties is null before check routine load properties
2. Change transactionStateChange reason to string
3. calculate current num by beId
4. Add kafka offset properties
5. Prefer to use previous be id
6. Add before commit listener of txn: if txn is committed after task is aborted, commit will be aborted
7. queryId of stream load plan = taskId
2019-04-28 10:33:50 +08:00
a2c3fb1507
Missing to set auth code ( #699 )
2019-04-28 10:33:50 +08:00
10cee6ecff
Add missing files ( #696 )
2019-04-28 10:33:50 +08:00
567d5de2de
Add a data consumer pool to reuse the data consumer ( #691 )
2019-04-28 10:33:50 +08:00
2314a3ecd4
Put begin txn into task scheduler ( #687 )
...
1. fix the nesting lock of db and txn
2. the txn of task will be init in task scheduler before take task from queue
2019-04-28 10:33:50 +08:00
20b2b2c37f
Modify interface ( #684 )
...
1. Add batch submit interface
2. Add Kafka Event callback to catch Kafka events
2019-04-28 10:33:50 +08:00
152606fbd6
Submit routine load task immediately ( #682 )
...
1. Use submit_routine_load_task instead of agentTaskQueue
2. Remove thrift dependency in StreamLoadPlanner and StreamLoadScanNode
2019-04-28 10:33:50 +08:00
9618d20a72
Add unit test ( #675 )
2019-04-28 10:33:50 +08:00
0820a29b8d
Implement the routine load process of Kafka on Backend ( #671 )
2019-04-28 10:33:50 +08:00
da308da17c
Fix bug that empty stream load return unexpected error msg ( #1052 )
2019-04-28 09:36:19 +08:00
88bd289caa
Modify thirdparties ( #1044 )
...
1. Update snappy from 1.1.4 to 1.1.7
2. disable ssl of librdkafka
2019-04-26 17:03:55 +08:00
642b2a3604
Add a java sample for stream load ( #1039 )
2019-04-26 13:30:50 +08:00
e8397d17f3
Change length of CHAR type when upon schema change ( #1035 )
...
CHAR type is of the fixed length string. Upon modify length
of CHAR type, the size of slice should be equal to new length.
2019-04-25 21:28:32 +08:00
101106d83e
Add left/right string function ( #1032 )
2019-04-25 17:29:34 +08:00
8031ab773b
Support add hll column in Schemachange ( #1033 )
2019-04-25 16:09:19 +08:00
3cc78fd952
Add trim in hdfs broker name node string ( #1028 )
2019-04-25 13:48:20 +08:00
566243c4e3
Fix unknown node id bug when stream load after decommission BE ( #993 )
2019-04-24 20:58:13 +08:00
95a06dcd2a
Change the buffer length for FloatToBuffer() method ( #1019 )
2019-04-24 20:11:06 +08:00
f852aa1cff
Support enable_insert_strict ( #1013 )
...
Add a variable enable_insert_strict, this default value is false. When
this value is set to true, insert will fail if there is any filtered
data. If this value is false, insert will ignore filtered data and
success
2019-04-24 17:06:53 +08:00
7933cc2209
Make add_pending_segment_group() idempotent ( #1001 )
2019-04-23 17:33:45 +08:00
a5494372b8
Fix some error in doc ( #998 )
2019-04-23 13:45:04 +08:00
e1a7773a03
Fix bug that failed to choose dest replica for version incomplete repair ( #996 )
...
We should skip the replica which version is larger than visible version.
2019-04-23 09:18:50 +08:00
84e3edebec
Fix bug that failed to get alive broker ( #989 )
2019-04-22 20:53:20 +08:00
22f93b5d7a
Fix doc in alter bloom filter ( #984 )
2019-04-22 14:07:12 +08:00
1cc2926a40
Fix wrong result in decimal type multiplication when carry is required ( #980 )
2019-04-21 17:12:58 +08:00
487c881092
Fix but that query cause be core when collecting query statistics ( #970 )
2019-04-19 17:08:49 +08:00
f273f59844
Refactor the code selecting rollup ( #947 )
2019-04-19 14:49:12 +08:00
22dc6119b9
Add some string functions doc ( #965 )
2019-04-19 09:50:06 +08:00
b9054f537b
Support UDAF ( #943 )
2019-04-17 19:08:37 +08:00
b99ddee4e6
Fix bug that HLL can be use as keys ( #937 )
2019-04-16 20:14:20 +08:00
b48a4ab6a0
Eliminate LOG(FATAL) when error occurs in ColumnData ( #935 )
2019-04-16 18:27:31 +08:00
ee62a0dd3e
Support adding columns to multiple rollup indexes in one ALTER TABLE stmt ( #931 )
2019-04-16 15:38:29 +08:00
cdd613c5a4
Add some string functions document ( #928 )
2019-04-16 10:14:49 +08:00
e5a5b6da16
Fix concat_ws return null when argument is null ( #923 )
...
#918
2019-04-15 19:54:02 +08:00
de1d1f715a
Add readme for new documentation framework ( #919 )
2019-04-15 11:01:26 +08:00
f2071d45a7
Change the download url for libevent ( #916 )
2019-04-13 22:20:09 +08:00
916c472e24
Fix bug in create table with BF ( #907 )
...
Fixed a bug in the Bloom filter where column names could not be found due to spaces
2019-04-11 16:32:33 +08:00
d3251a19f7
Modify the method to obtain some metrics ( #904 )
2019-04-10 19:37:48 +08:00
c1b8c11dbb
Fix null pointer error when async load ( #902 )
...
1. Add parallel number property in fragment
2. Parallel number will be init in contrustor of fragment
3. Async plan will use the default value while sync plan will use the parallelExecInstanceNum of SessionVariable
2019-04-10 16:00:12 +08:00
9c1de6ce38
Fix HLL compaction bug. ( #901 )
...
1. Cumulative Compaction in HLL will core dump because of null pointer
2019-04-10 10:37:23 +08:00
2a3bf5842f
Parallel fragment exec instance ( #851 )
2019-04-09 10:05:39 +08:00
29f6665826
Change value pass to value reference ( #884 )
2019-04-08 10:44:23 +08:00
92f17afb69
Fix a problem which may cause nullptr in ResultBufferMgr ( #882 )
...
Because there is no lock in the middle call to findBlock(),
even if the first call returns non-null, the second call
may still get a null block
2019-04-06 14:18:34 +08:00
61e356b8a5
Fix docs 'CANCEL DECOMMISSION' ( #879 )
2019-04-04 17:38:29 +08:00
d9f93e345e
Revert "Fix issue #866 : insert stmt not merge commit info correctly ( #868 )" ( #880 )
...
This reverts commit d5a6d1cc6f1eb4d297a8cbc7cbbe02544a2c402c.
2019-04-04 17:33:55 +08:00
c0fbc84381
Fix bug that ScanBytes is when collect executing query's infos ( #869 )
2019-04-03 18:27:50 +08:00
