Doris should provide a http api to return backends list for connectors to submit stream load,
and without privilege checking, which can let common user to use it
1. By default , Spark connector must write all fields value to `Doris` table .
In this feature , user can specify part of fields to write , even specify the order of the fields to write.
eg:
I have a table named `student` which has three columns (name,gender,age) ,
creating table sql as following:
```sql
create table student (name varchar(255), gender varchar(10), age int) duplicate key (name) distributed by hash(name) buckets 2;
```
Now , I just want to write values to two columns : name , gender.
The code as following:
```scala
val df = spark.createDataFrame(Seq(
("m", "zhangsan"),
("f", "lisi"),
("m", "wangwu")
))
df.write
.format("doris")
.option("doris.fenodes", dorisFeNodes)
.option("doris.table.identifier", dorisTable)
.option("user", dorisUser)
.option("password", dorisPwd)
//specify your fields or the order
.option("doris.write.field", "gender,name")
.save()
```
schema change fail as memory allocation fail on row block sorting.
however, it should do internal sorting first before schema change fail
as memory allocation fail on row block sorting in case there are enough
memory after internal sorting.
## Case
In the load process, each tablet will have a memtable to save the incoming data,
and if the data in a memtable is larger than 100MB, it will be flushed to disk as a `segment` file. And then
a new memtable will be created to save the following data/
Assume that this is a table with N buckets(tablets). So the max size of all memtables will be `N * 100MB`.
If N is large, it will cost too much memory.
So for memory limit purpose, when the size of all memtables reach a threshold(2GB as default), Doris will
try to flush all current memtables to disk(even if their size are not reach 100MB).
So you will see that the memtable will be flushed when it's size reach `2GB/N`, which maybe much smaller
than 100MB, resulting in too many small segment files.
## Solution
When decide to flush memtable to reduce memory consumption, NOT to flush all memtable, but to flush part
of them.
For example, there are 50 tablets(with 50 memtables). The memory limit is 1GB, so when each memtable reach
20MB, the total size reach 1GB, and flush will occur.
If I only flush 25 of 50 memtables, then next time when the total size reach 1GB, there will be 25 memtables with
size 10MB, and other 25 memtables with size 30MB. So I can flush those memtables with size 30MB, which is larger
than 20MB.
The main idea is to introduce some jitter during flush to ensure the small unevenness of each memtable, so as to ensure that flush will only be triggered when the memtable is large enough.
In my test, loading a table with 48 buckets, mem limit 2G, in previous version, the average memtable size is 44MB,
after modification, the average size is 82MB
Add a use_path_style property for S3
Upgrade hadoop-common and hadoop-aws to 2.8.0 to support path style property
Fix some S3 URI bugs
Add some logs for tracing load process.
1. optimize error message when using batch delete
2. rename session variable is_report_success to enable_profile
3. add table name to OlapScanner profile
when be has excepiton, fe doesn't log the BackendHbResponse info, so we can't know which be has exception
the exception log is：
`WARN (heartbeat mgr|31) [HeartbeatMgr.runAfterCatalogReady():141] get bad heartbeat response: type: BACKEND, status: BAD, msg: java.net.ConnectException: Connection refused (Connection refused)
`
so need add toString(), then fe can log the BackendHbResponse info
This CL mainly changes:
1. Add star schema benchmark tools in `tools/ssb-tools`, for user to easy load and test with SSB data set.
2. Disable the segment cache for some read scenario such as compaction and alter operation.(Fix#6924 )
3. Fix a bug that `max_segment_num_per_rowset` won't work(Fix#6926)
4. Enable `enable_batch_delete_by_default` by default.
1. Simplify the use of flink connector like other stream sink by GenericDorisSinkFunction.
2. Add the use cases of flink connector.
## Use case
```
env.fromElements("{\"longitude\": \"116.405419\", \"city\": \"北京\", \"latitude\": \"39.916927\"}")
.addSink(
DorisSink.sink(
DorisOptions.builder()
.setFenodes("FE_IP:8030")
.setTableIdentifier("db.table")
.setUsername("root")
.setPassword("").build()
));
```
When loading meta by meta_tool goes wrong, we only get an error code from `json2pb`,
which is inconvenient for us to locate the problem.
This change is adding error message when loading meta goes wrong.
Log change is like below.
```
# before
./meta_tool --root_path=/home/disk1/qjl/mydoris/be/storage --operation=load_meta --json_meta_path=/home/disk1/qjl/data/meta-json.json
WARNING: Logging before InitGoogleLogging() is written to STDERR
I1020 11:41:56.564241 74937 data_dir.cpp:837] path: /home/disk1/qjl/mydoris/be/storage total capacity: 7750843404288, available capacity: 7583325925376
I1020 11:41:56.564415 74937 data_dir.cpp:275] path: /home/disk1/qjl/mydoris/be/storage, hash: 7528840506668047470
load meta failed, status:-1410
# after
./meta_tool --root_path=/home/disk1/qjl/mydoris/be/storage --operation=load_meta --json_meta_path=/home/disk1/qjl/data/meta-json.json
WARNING: Logging before InitGoogleLogging() is written to STDERR
I1020 14:41:40.084342 50727 data_dir.cpp:837] path: /home/disk1/qjl/mydoris/be/storage total capacity: 7750843404288, available capacity: 7584601022464
I1020 14:41:40.084496 50727 data_dir.cpp:275] path: /home/disk1/qjl/mydoris/be/storage, hash: 7528840506668047470
E1020 14:41:40.163007 50727 tablet_meta_manager.cpp:161] JSON to protobuf message failed: Fail to decode base64 string=0
load meta failed, status:-1410
```
String.valueOf() returns string "null" with null input, in which case requests with no SQL
will be accepted by TableQueryPlanAction unexpectedly with potential risk.
