database/doris
2
0
Fork 0
Code Issues Pull Requests Actions 3 Packages Projects Releases Wiki Activity

[fix](memory) query check cancel while waiting for memory in Allocator, and optimize log (#19967)

Browse Source 
After the query check process memory exceed limit in Allocator, it will wait up to 5s.
Before, Allocator will not check whether the query is canceled while waiting for memory, this causes the query to not end quickly.
This commit is contained in:
Xinyi Zou
2023-05-24 11:08:48 +08:00
committed by GitHub
parent 08ec5e2eb5
commit cf7a74f6ec
5 changed files with 46 additions and 27 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
be/src/runtime/thread_context.cpp
View File

@ -20,7 +20,6 @@
#include "common/signal_handler.h"
#include "runtime/runtime_state.h"
#include "util/doris_metrics.h" // IWYU pragma: keep
#include "util/uid_util.h"
namespace doris {
class MemTracker;
@ -33,15 +32,14 @@ ThreadContextPtr::ThreadContextPtr() {
}
AttachTask::AttachTask(const std::shared_ptr<MemTrackerLimiter>& mem_tracker,
const std::string& task_id, const TUniqueId& fragment_instance_id) {
const TUniqueId& task_id, const TUniqueId& fragment_instance_id) {
thread_context()->attach_task(task_id, fragment_instance_id, mem_tracker);
}
AttachTask::AttachTask(RuntimeState* runtime_state) {
doris::signal::query_id_hi = runtime_state->query_id().hi;
doris::signal::query_id_lo = runtime_state->query_id().lo;
thread_context()->attach_task(print_id(runtime_state->query_id()),
runtime_state->fragment_instance_id(),
thread_context()->attach_task(runtime_state->query_id(), runtime_state->fragment_instance_id(),
runtime_state->query_mem_tracker());
}