Support merge IN predicate when exist remote target(e.g. shuffle hash join).
Remote the code that IN predicate implicit conversion to Bloom filter then exist remote target.
Close related #7546
Add a new FE config `repair_slow_replica`
when this config is true, Doris will try to delete the replica
with the largest number of versions, and then rebalance the replica.
Usually, when the number of versions of a certain replica is much higher
then that of other replicas, there are some problems with the current be's compilation.
Migrating to other machines can typically solve this problem.
Add partitionNum, tabletNum, cardinality in SqlBlockRule to block large/slow sql.
1. set partitionNum, tabletNum, cardinality as limitations to block sqls
2. compatible with lower version
3. add unit tests
4. add docs
This pr is for #7096 , which is add a rewrite rule for infer predicate.
For example:
origin stmt: select * from t1, t2, t3 where t1.id=t2.id and t2.i=t3.id and t2.id = 1
rewrite stmt: select * from t1, t2, t3 where t1.id=t2.id and t2.i=t3.id and t2.id = 1 and t1.id=1 and t3.id=1
+ Add a switch enable_infer_predicate to control whether to perform predicate expansion.
+ Register a new rule InferFiltersrule and add it to GlobalState.
+ Traverse Conjunct to construct on/where equivalence connection, numerical connection and isNullPredicate.
+ Infer all equivalence connections
+ Construct additional numerical connections and isNullPredicate
1. Change `brpc_socket_max_unwritten_bytes` to 1GB
This can make the system more fault-tolerant.
Especially in the case of high system load, try to reduce EOVERCROWDED errors.
2. Change `brpc_max_body_size` to 3GB
To handle some large object such as bitmap or string.
1. Refactor the scheduling logic of broker load. Details see #7367
2. Fix bug that loadedBytes in SHOW LOAD result is wrong.
3. Cancel the thread of LoadTimeoutChecker
Now for PENDING load jobs, there will be no timeout. And the timeout of a load job
start when pending load task is scheduled.
4. Fix a bug that the loading task is never submitted to the pool.
The logic of BlockedPolicy is wrong. We should make sure the task is submitted to the pool,
or the RejectedExecutionException should be thrown.
5. Now the transaction of a load job will begin in pending task, instead of when submitting the job.
At present, there are defects in the chapter on debugging FE in doc. My colleagues and I stepped on the pit when
building the debugging environment, so I want to improve this chapter in combination with my own stepping on the pit
experience.
The following is my explanation of the changes:
1. mkdir -p ./thirdparty/installed/bin
explain: When I downloaded versions 0.14 and 0.15, there were no files under thirdparty, so I didn't know whether to
create it myself or what to do. Finally, I decided to create it myself. I think it's necessary to add instructions here.
2. Add installation thrift@0.13.0 Failed handling method.
explain: My colleagues and I failed to find the installation package when executing the installation command, and finally
found a solution on GitHub. Therefore, I added the handling method of the problem to avoid other Mac users from
getting stuck in this place.
3. Fixed an error in the generated code description.
explain: Before I finished building the code, I debugged FE, and I failed all the time. Idea hints that no files can be found.
Later, after consulting with morningman in wechat group, it was understood that `mvn install -DskipTests` does not
need to execute `mvn generate-sources` after execution. This is inconsistent with the description in the document and
needs to be corrected.
Close related #7334
1. Fix bug describe in [Bug] show frontends cause FE oom #7334
2. Fix error of CurrentConnected fields in show frontends result.
3. Add more FAQ
Add `sink.batch.size` `sink.max-retries` options in `Doris Spark-connector`.
Be consistent with `link-connector` options .
eg:
```scala
df.write
.format("doris")
// specify maximum number of lines in a single flushing
.option("sink.batch.size",2048)
// specify number of retries after writing failed
.option("sink.max-retries",3)
.save()
```
Transfer RowBatch in Protobuf Request to Controller Attachment,
when the maximum length of the RowBatch in the Protobuf Request is exceeded.
This can avoid reaching the upper limit of the Protobuf Request length (2G),
and it is expected that performance can be improved.
1. The clang format action will be triggered when a PR is submitted.
2. Skywalking eyes actions will be triggered when a PR is submitted and after merging to master branch.
