Current cumulative point calculate algorithm may skip singleton rowset when the rowset has only one segment and with NONOVERLAPPING flag. When a tablet is new created and cumulate many singleton rowsets, cumulative point will be calculated as the max version + 1, and then cumulative compaction couldn't pick any rowsets and compaction failed, and
will lead the next base compaction on this tablet with all rowsets, which can also cause memory consume problem, suppose there are thousands of rowsets.
All singleton rowsets must be newly wrote by delta writer and hasn't
do any compaction, we should place cumulative point before any of these rowsets.
Fixes#3706
DataSink uses instance and query MemTracker from RuntimeState, therefore it should be destructed before RuntimeState. Otherwise memory corruption and segfault could happen.
```
CREATE TABLE `query_detail` (
`query_id` varchar(100) NULL COMMENT "",
`start_time` datetime NULL COMMENT "",
`end_time` datetime NULL COMMENT "",
`latency` int(11) NULL COMMENT "unit is milliseconds",
`state` varchar(20) NULL COMMENT "RUNNING/FINISHED/FAILED",
`sql` varchar(1024) NULL COMMENT ""
)
DUPLICATE KEY(`query_id`)
SELECT COUNT(*) FROM query_detail WHERE start_time >= '2020-05-27 14:52:16' AND start_time < '2020-05-27 14:52:31';
```
The above query will core because of ZoneMap only in query_id.
Use start_time to match ZoneMap cause this core.
Add a JSON format for existing metrics like this.
```
{
"tags":
{
"metric":"thread_pool",
"name":"thrift-server-pool",
"type":"active_thread_num"
},
"unit":"number",
"value":3
}
```
I add a new JsonMetricVisitor to handle the transformation.
It's not to modify existing PrometheusMetricVisitor and SimpleCoreMetricVisitor.
Also I add
1. A unit item to indicate the metric better
2. Cloning tablet statistics divided by database.
3. Use white space to replace newline in audit.log
Change the load label of audit plugin as:
`audit_yyyyMMdd_HHmmss_feIdentity`.
The `feIdentity` is got from the FE which run this plugin, currently just use FE's IP_editlog_port.
This CL mainly changes:
1. Support `SELECT INTO OUTFILE` command.
2. Support export query result to a file via Broker.
3. Support CSV export format with specified column separator and line delimiter.
* Support bitmap_intersect
Support aggregate function Bitmap Intersect, it is mainly used to take intersection of grouped data.
The function 'bitmap_intersect(expr)' calculates the intersection of bitmap columns and returns a bitmap object.
The defination is following:
FunctionName: bitmap_intersect,
InputType: bitmap,
OutputType: bitmap
The scenario is as follows:
Query which users satisfy the three tags a, b, and c at the same time.
```
select bitmap_to_string(bitmap_intersect(user_id)) from
(
select bitmap_union(user_id) user_id from bitmap_intersect_test
where tag in ('a', 'b', 'c')
group by tag
) a
```
Closed#3552.
* Add docs of bitmap_union and bitmap_intersect
* Support null of bitmap_intersect
Why this case happened
In current implement, translation into dsl only if it is not the first charactor.
Thus, when sql is write like '%abc', translation would not run.
How fixed
Now, translation will trigger with charactor '?' or '*'
if it is the first charactor, translate directly
else, check the preceding char is escaped or not to determin translation or not
Ref https://github.com/apache/incubator-doris/issues/3566
Introduce trace utility from Kudu to BE. This utility has been widely used in Kudu,
Impala also import this trace utility.
This trace util is used for tracing each phases in a thread, and can be dumped to
string to see each phases' time cost and diagnose which phase cost more time.
This util store a Trace object as a threadlocal variable, we can add trace entries
which record the current file name, line number, user specified symbols and
timestamp to this object, and it's able to add some counters to this Trace
object. And then, it can be dumped to human readable string.
There are some helpful macros defined in trace.h, here is a simple example for
usage:
```
scoped_refptr<Trace> t1(new Trace); // New 2 traces
scoped_refptr<Trace> t2(new Trace);
t1->AddChildTrace("child_trace", t2.get()); // t1 add t2 as a child named "child_trace"
TRACE_TO(t1, "step $0", 1); // Explicitly trace to t1
usleep(10);
// ... do some work
ADOPT_TRACE(t1.get()); // Explicitly adopt to trace to t1
TRACE("step $0", 2); // Implicitly trace to t1
{
// The time spent in this scope is added to counter t1.scope_time_cost
TRACE_COUNTER_SCOPE_LATENCY_US("scope_time_cost");
ADOPT_TRACE(t2.get()); // Adopt to trace to t2 for the duration of the current scope
TRACE("sub start"); // Implicitly trace to t2
usleep(10);
// ... do some work
TRACE("sub before loop");
for (int i = 0; i < 10; ++i) {
TRACE_COUNTER_INCREMENT("iterate_count", 1); // Increase counter t2.iterate_count
MicrosecondsInt64 start_time = GetMonoTimeMicros();
usleep(10);
// ... do some work
MicrosecondsInt64 end_time = GetMonoTimeMicros();
int64_t dur = end_time - start_time;
// t2's simple histogram metric with name prefixed with "lbm_writes"
const char* counter = BUCKETED_COUNTER_NAME("lbm_writes", dur);
TRACE_COUNTER_INCREMENT(counter, 1);
}
TRACE("sub after loop");
}
TRACE("goodbye $0", "cruel world"); // Automatically restore to trace to t1
std::cout << t1->DumpToString(Trace::INCLUDE_ALL) << std::endl;
```
output looks like:
```
0514 02:16:07.988054 (+ 0us) trace_test.cpp:76] step 1
0514 02:16:07.988112 (+ 58us) trace_test.cpp:80] step 2
0514 02:16:07.988863 (+ 751us) trace_test.cpp:103] goodbye cruel world
Related trace 'child_trace':
0514 02:16:07.988120 (+ 0us) trace_test.cpp:85] sub start
0514 02:16:07.988188 (+ 68us) trace_test.cpp:88] sub before loop
0514 02:16:07.988850 (+ 662us) trace_test.cpp:101] sub after loop
Metrics: {"scope_time_cost":744,"child_traces":[["child_trace",{"iterate_count":10,"lbm_writes_lt_1ms":10}]]}
```
Exclude the original source code, this patch
do the following work to adapt to Doris:
- Rename "kudu" namespace to "doris"
- Update some names to the existing function names in Doris, i.g. strings::internal::SubstituteArg::kNoArg -> strings::internal::SubstituteArg::NoArg
- Use doris::SpinLock instead of kudu::simple_spinlock which hasn't been imported
- Use manual malloc() and free() instead of kudu::Arena which hasn't been imported
- Use manual rapidjson::Writer instead of kudu::JsonWriter which hasn't been imported
- Remove all TRACE_EVENT related unit tests since TRACE_EVENT is not imported this time
- Update CMakeLists.txt
NOTICE(#3622):
This is a "revert of revert pull request".
This pr is mainly used to synthesize the PRs whose commits were
scattered and submitted due to the wrong merge method into a complete single commit.
Ref https://github.com/apache/incubator-doris/issues/3566
Introduce trace utility from Kudu to BE. This utility has been widely used in Kudu,
Impala also import this trace utility.
This trace util is used for tracing each phases in a thread, and can be dumped to
string to see each phases' time cost and diagnose which phase cost more time.
This util store a Trace object as a threadlocal variable, we can add trace entries
which record the current file name, line number, user specified symbols and
timestamp to this object, and it's able to add some counters to this Trace
object. And then, it can be dumped to human readable string.
There are some helpful macros defined in trace.h, here is a simple example for
usage:
```
scoped_refptr<Trace> t1(new Trace); // New 2 traces
scoped_refptr<Trace> t2(new Trace);
t1->AddChildTrace("child_trace", t2.get()); // t1 add t2 as a child named "child_trace"
TRACE_TO(t1, "step $0", 1); // Explicitly trace to t1
usleep(10);
// ... do some work
ADOPT_TRACE(t1.get()); // Explicitly adopt to trace to t1
TRACE("step $0", 2); // Implicitly trace to t1
{
// The time spent in this scope is added to counter t1.scope_time_cost
TRACE_COUNTER_SCOPE_LATENCY_US("scope_time_cost");
ADOPT_TRACE(t2.get()); // Adopt to trace to t2 for the duration of the current scope
TRACE("sub start"); // Implicitly trace to t2
usleep(10);
// ... do some work
TRACE("sub before loop");
for (int i = 0; i < 10; ++i) {
TRACE_COUNTER_INCREMENT("iterate_count", 1); // Increase counter t2.iterate_count
MicrosecondsInt64 start_time = GetMonoTimeMicros();
usleep(10);
// ... do some work
MicrosecondsInt64 end_time = GetMonoTimeMicros();
int64_t dur = end_time - start_time;
// t2's simple histogram metric with name prefixed with "lbm_writes"
const char* counter = BUCKETED_COUNTER_NAME("lbm_writes", dur);
TRACE_COUNTER_INCREMENT(counter, 1);
}
TRACE("sub after loop");
}
TRACE("goodbye $0", "cruel world"); // Automatically restore to trace to t1
std::cout << t1->DumpToString(Trace::INCLUDE_ALL) << std::endl;
```
output looks like:
```
0514 02:16:07.988054 (+ 0us) trace_test.cpp:76] step 1
0514 02:16:07.988112 (+ 58us) trace_test.cpp:80] step 2
0514 02:16:07.988863 (+ 751us) trace_test.cpp:103] goodbye cruel world
Related trace 'child_trace':
0514 02:16:07.988120 (+ 0us) trace_test.cpp:85] sub start
0514 02:16:07.988188 (+ 68us) trace_test.cpp:88] sub before loop
0514 02:16:07.988850 (+ 662us) trace_test.cpp:101] sub after loop
Metrics: {"scope_time_cost":744,"child_traces":[["child_trace",{"iterate_count":10,"lbm_writes_lt_1ms":10}]]}
```
Exclude the original source code, this patch
do the following work to adapt to Doris:
- Rename "kudu" namespace to "doris"
- Update some names to the existing function names in Doris, i.g. strings::internal::SubstituteArg::kNoArg -> strings::internal::SubstituteArg::NoArg
- Use doris::SpinLock instead of kudu::simple_spinlock which hasn't been imported
- Use manual malloc() and free() instead of kudu::Arena which hasn't been imported
- Use manual rapidjson::Writer instead of kudu::JsonWriter which hasn't been imported
- Remove all TRACE_EVENT related unit tests since TRACE_EVENT is not imported this time
- Update CMakeLists.txt
1. Delete Invalid Counter In Data_Stream_Sender. (#3598)
2. Add Counter For PartitionHashTable of PartitionAggregationNode:
* Hash Probe Method
* Row processed by Aggregation
* HashFilledBuckets: Counter How Many FilledBuckets in Aggragation
* HTResize: Counter How Many Resize of HashTable
* HashProbe: Counter Probe of HashTable
* HashFailedProbe: Counter Failed Probe of HashTable
* HashTravelLength: Total TravelLength for Probe
* HashCollisions: Counter of HashCollision
3. Del some unecessary code in PartitionHashTable by template
#3479
Here I try to explain the cause of the problem and how to fix it.
**The Cause of The problem**
Take the case in issue(#3479 ) as an example:
The general results are as follows:
```
GET table/_doc/_search
{"query":{"match_all":{}},"stored_fields":"_none_","docvalue_fields":["k1"],"sort":["_doc"],"size":100}
{
"took": 6,
"timed_out": false,
"_shards": {
……
},
"hits": {
"total": 3,
"max_score": null,
"hits": [
{
"_index": "table",
"_score": null,
"sort": [
0
]
},
{
"_index": "table",
"_score": null,
"fields": {
"k1": [
"kkk1"
]
},
"sort": [
0
]
},
{
"_index": "table",
"_score": null,
"sort": [
0
]
}
]
}
}
```
But in Doris on ES，Be fetched data parallelly on all shards, and use `filter_path` to reduce the network cost. The process will be as follows:
```
GET table/_doc/_search?preference=_shards:1&filter_path=_scroll_id,hits.hits._source,hits.total,_id,hits.hits._source.fields,hits.hits.fields
{"query":{"match_all":{}},"stored_fields":"_none_","docvalue_fields":["k1"],"sort":["_doc"],"size":100}
{
"hits": {
"total": 0
}
}
GET table/_doc/_search?preference=_shards:2&filter_path=_scroll_id,hits.hits._source,hits.total,_id,hits.hits._source.fields,hits.hits.fields
{"query":{"match_all":{}},"stored_fields":"_none_","docvalue_fields":["k1"],"sort":["_doc"],"size":100}
{
"hits": {
"total": 1
}
}
GET table/_doc/_search?preference=_shards:3&filter_path=_scroll_id,hits.hits._source,hits.total,_id,hits.hits._source.fields,hits.hits.fields
{"query":{"match_all":{}},"stored_fields":"_none_","docvalue_fields":["k1"],"sort":["_doc"],"size":100}
{
"hits": {
"total": 1,
"hits": [
{
"fields": {
"k1": [
"kkk1"
]
}
}
]
}
}
```
*Scan-Worker On BE which processed result of shard2 will failed.*
**The reasons are as follows:**
1. "filter_path" causes the hits.hits object not exist.
2. In the current implementation, if there are some data rows（total > 0）, the hits.hits. object must be an array
**How To Fix it**
Two Method:
1. modify "filter_path" to contain the hits.
Pros: Fixed Code is very simple
Cons: More network cost
2. Deal with the case where fields are missing in a batch.
Pros: No loss of performance
Cons: Code is more complex
Performance first, I use Method2.
**Design**
1. Add a variable "_doc_value_mode" into Class "EsScrollParser" to =indicate whether the data processed by this parser is doc_value_mode or not.
2. "_doc_value_mode" is passed from ESScollReader <- ESScanner <- ScrollQueryBuilder::build() that determines whether DSL is enable doc_value_mode
3. When hits.hits of response from ES is empty and total > 0. We know there are data lines, but the corresponding fields do not exist. EsScrollParser will use "_doc_value_mode" and _total to construct _total lines which fields are assigned with 'NULL'
LSAN detected errors have been fixed by a prior pathch (#3326), but
there are still some ASAN detected errors.
This patch try to fix these errors to make Doris BE more robustness.
And then we can add CI run in LSAN/ASAN mode to detect memory errors
as early as possible.
Each test case in ExternalScanContextMgrTest may cost 1 minitue
which is too long, we'd better disable backgrounp scan context
gc to speed up unit test.
When BE sets `ignore_broken_disk` to true, it's expected that non-exist path in storage_root_path won't prevent BE from launching, but in 0.12 BE fails to launch in such scenario.
```
W0506 14:46:11.039953 17040 options.cpp:64] path can not be canonicalized. may be not exist. path=/data11/olap
W0506 14:46:11.040014 17040 options.cpp:141] failed to parse store path /data11/olap, res=-203
```
The reason is that #2861 adds a path existence check in `parse_root_path` which precedes the usage of `ignore_broken_disk` in the main method.
ImplementaItion Notes
NodeChannel
_cur_batch -> _pending_batches: when _cur_batch is filled up, move it to _pending_batches.
add_row() just produce batches.
try_send_and_fetch_status() tries to consume one pending batch. If has in flight packet, skip send in this round.
So we can add one sender thread to be in charge of all node channels try_send.
IndexChannel
init(), open() stay the same.
Use for_each_node_channel() to expose the detailed changes of NodeChannel.(It's more easy to read & modify)
Sender thread
See func OlapTableSink::_send_batch_process()
Why use polling？
If we use wait/notify, it will notify when generate a new batch. We can't skip sending this batch, coz it won't notify the same batch again. So wait/notify can't avoid blocking simply.
So I choose polling.
It's wasting to continuously try_send(), but it's difficult to set the suitable polling interval. Thus, I add std::this_thread::yield() to give up the time slice, give priority to other process/threads (if there are other process/threads waiting in the queue).
RandomAccessFileOptions, WritableFileOptions, RandomRWFileOptions
defined as a struct but previously declared as a class; this is valid,
but will result in compile warning or error under clang compiler
