In the following case, data inconsistency would happen between multiple replicas
current delta writer only writes a few lines of data (which meas the write() method only called once)
writer failed when init()(which is called at the fist time we call write()), and current tablet is recorded in _broken_tablets
delta writer closed, and in the close() method, delta writer found it's not inited, treat such case as an empty load, it will try to init again, which would create an empty rowset.
tablet sink received the error report in rpc response, marked the replica as failed, but since the quorum replicas are succeed, so the following load commit operation will succeed.
FE send publish version task to each be, the one with empty rowset will publish version successfully.
We got 2 replica with data and 1 empty replica.
Add conf enable_query_memroy_overcommit
If true, when the process does not exceed the soft mem limit, the query memory will not be limited; when the process memory exceeds the soft mem limit, the query with the largest ratio between the currently used memory and the exec_mem_limit will be canceled.
If false, cancel query when the memory used exceeds exec_mem_limit, same as before.
When the system MemAvailable is less than the warning water mark, or the memory used by the BE process exceeds the mem soft limit, run minor gc and try to release cache.
When the MemAvailable of the system is less than the low water mark, or the memory used by the BE process exceeds the mem limit, run fucc gc, try to release the cache, and start canceling from the query with the largest memory usage until the memory of mem_limit * 20% is released.
1.Support in bitmap syntax, like 'where k1 in (select bitmap_column from tbl)';
2.Support bitmap runtime filter. Generate a bitmap filter using the right table bitmap and push it down to the left table storage layer for filtering.
This PR is to make sharing hash table for broadcast more robust:
Add a session variable to enable/disable this function.
Do not block the hash join node's close function.
Use shared pointer to share hash table and runtime filter in broadcast join nodes.
The Hash join node that doesn't need to build the hash table will close the right child without reading any data(the child will close the corresponding sender).
Fix Thread pool token was shut down error.
This is because when there are more than 1 fragment of a query on one BE, the thread token maybe
reset incorrectly, causing thread token shutdown earlier.
cherry-pick from master
Introduced from #13021
boost::stacktrace::stacktrace() has memory leak, so use glog internal func to print stacktrace.
The reason for the memory leak of boost::stacktrace is that a state is saved in the thread local of each thread but not actively released. The test found that each thread leaked about 100M after calling boost::stacktrace.
refer to:
boostorg/stacktrace#118boostorg/stacktrace#111
mem tracker can be logically divided into 4 layers: 1)process 2)type 3)query/load/compation task etc. 4)exec node etc.
type includes
enum Type {
GLOBAL = 0, // Life cycle is the same as the process, e.g. Cache and default Orphan
QUERY = 1, // Count the memory consumption of all Query tasks.
LOAD = 2, // Count the memory consumption of all Load tasks.
COMPACTION = 3, // Count the memory consumption of all Base and Cumulative tasks.
SCHEMA_CHANGE = 4, // Count the memory consumption of all SchemaChange tasks.
CLONE = 5, // Count the memory consumption of all EngineCloneTask. Note: Memory that does not contain make/release snapshots.
BATCHLOAD = 6, // Count the memory consumption of all EngineBatchLoadTask.
CONSISTENCY = 7 // Count the memory consumption of all EngineChecksumTask.
}
Object pointers are no longer saved between each layer, and the values of process and each type are periodically aggregated.
other fix:
In [fix](memtracker) Fix transmit_tracker null pointer because phamp is not thread safe #13528, I tried to separate the memory that was manually abandoned in the query from the orphan mem tracker. But in the actual test, the accuracy of this part of the memory cannot be guaranteed, so put it back to the orphan mem tracker again.
When the load mem hard limit reached, all load channel should wait on the lock of LoadChannelMgr, util current reduce mem work finished. In current implementation, there's a bug might cause some threads be woke up before reduce mem work finished:
thread A found that soft limit reached, picked a load channel and waiting for reduce memory work finish.
The memory keep increasing
thread B found that hard limit reached (either the load mem hard limit, or process soft limit), it picked a load channel to reduce memory and set the variable _should_wait_flush to true
thread C found that _should_wait_flush is true, waiting on _wait_flush_cond
thread A finished it's reduce memory work, found that _should_wait_flush is true, set it to false, and notify all threads.
thread C is woke up and pick a load channel to do the reduce memory work, and now thread B's work is not finished.
We can see 2 threads doing reduce memory work when hard limit reached, it's quite confusing.
