tursom/memgraph
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Improve concurrency model

Browse Source
This commit is contained in:
Andi Skrgat 2024-01-25 15:20:14 +01:00
parent 62d1b68c2f
commit 9bc0d9425e
3 changed files with 14 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
src/coordination/CMakeLists.txt
View File

@ -25,5 +25,5 @@ target_sources(mg-coordination
target_include_directories(mg-coordination PUBLIC include)
target_link_libraries(mg-coordination
PUBLIC mg::utils mg::rpc mg::slk mg::io mg::repl_coord_glue rangev3
PUBLIC mg::utils mg::rpc mg::slk mg::io mg::repl_coord_glue lib::rangev3
)

20
src/coordination/coordinator_data.cpp
View File

@ -22,8 +22,6 @@ namespace memgraph::coordination {
CoordinatorData::CoordinatorData() {
auto find_instance = [](CoordinatorData *coord_data, std::string_view instance_name) -> CoordinatorInstance & {
std::shared_lock<utils::RWLock> lock{coord_data->coord_data_lock_};
auto instance = std::ranges::find_if(
coord_data->registered_instances_,
[instance_name](const CoordinatorInstance &instance) { return instance.InstanceName() == instance_name; });
@ -34,6 +32,7 @@ CoordinatorData::CoordinatorData() {
};
replica_succ_cb_ = [find_instance](CoordinatorData *coord_data, std::string_view instance_name) -> void {
auto lock = std::lock_guard{coord_data->coord_data_lock_};
spdlog::trace("Instance {} performing replica successful callback", instance_name);
auto &instance = find_instance(coord_data, instance_name);
MG_ASSERT(instance.IsReplica(), "Instance {} is not a replica!", instance_name);
@ -41,6 +40,7 @@ CoordinatorData::CoordinatorData() {
};
replica_fail_cb_ = [find_instance](CoordinatorData *coord_data, std::string_view instance_name) -> void {
auto lock = std::lock_guard{coord_data->coord_data_lock_};
spdlog::trace("Instance {} performing replica failure callback", instance_name);
auto &instance = find_instance(coord_data, instance_name);
MG_ASSERT(instance.IsReplica(), "Instance {} is not a replica!", instance_name);
@ -48,6 +48,7 @@ CoordinatorData::CoordinatorData() {
};
main_succ_cb_ = [find_instance](CoordinatorData *coord_data, std::string_view instance_name) -> void {
auto lock = std::lock_guard{coord_data->coord_data_lock_};
spdlog::trace("Instance {} performing main successful callback", instance_name);
auto &instance = find_instance(coord_data, instance_name);
MG_ASSERT(instance.IsMain(), "Instance {} is not a main!", instance_name);
@ -55,6 +56,7 @@ CoordinatorData::CoordinatorData() {
};
main_fail_cb_ = [this, find_instance](CoordinatorData *coord_data, std::string_view instance_name) -> void {
auto lock = std::lock_guard{coord_data->coord_data_lock_};
spdlog::trace("Instance {} performing main failure callback", instance_name);
auto &instance = find_instance(coord_data, instance_name);
MG_ASSERT(instance.IsMain(), "Instance {} is not a main!", instance_name);
@ -80,7 +82,6 @@ CoordinatorData::CoordinatorData() {
auto CoordinatorData::DoFailover() -> DoFailoverStatus {
using ReplicationClientInfo = CoordinatorClientConfig::ReplicationClientInfo;
std::lock_guard<utils::RWLock> lock{coord_data_lock_};
auto replica_instances = registered_instances_ | ranges::views::filter(&CoordinatorInstance::IsReplica);
@ -94,10 +95,10 @@ auto CoordinatorData::DoFailover() -> DoFailoverStatus {
std::vector<ReplicationClientInfo> repl_clients_info;
repl_clients_info.reserve(std::ranges::distance(replica_instances));
auto not_chosen_replica_instance = [&chosen_replica_instance](const CoordinatorInstance &instance) {
auto const not_chosen_replica_instance = [&chosen_replica_instance](const CoordinatorInstance &instance) {
return instance != *chosen_replica_instance;
};
auto not_main = [](const CoordinatorInstance &instance) { return !instance.IsMain(); };
auto const not_main = [](const CoordinatorInstance &instance) { return !instance.IsMain(); };
// TODO (antoniofilipovic): Should we send also data on old MAIN???
// TODO: (andi) Don't send replicas which aren't alive
@ -117,7 +118,7 @@ auto CoordinatorData::DoFailover() -> DoFailoverStatus {
}
auto CoordinatorData::ShowMain() const -> std::optional<CoordinatorInstanceStatus> {
std::shared_lock<utils::RWLock> lock{coord_data_lock_};
auto lock = std::shared_lock{coord_data_lock_};
auto main_instance = std::ranges::find_if(registered_instances_, &CoordinatorInstance::IsMain);
if (main_instance == registered_instances_.end()) {
return std::nullopt;
@ -129,7 +130,7 @@ auto CoordinatorData::ShowMain() const -> std::optional<CoordinatorInstanceStatu
};
auto CoordinatorData::ShowReplicas() const -> std::vector<CoordinatorInstanceStatus> {
std::shared_lock<utils::RWLock> lock{coord_data_lock_};
auto lock = std::shared_lock{coord_data_lock_};
std::vector<CoordinatorInstanceStatus> instances_status;
for (const auto &replica_instance : registered_instances_ | ranges::views::filter(&CoordinatorInstance::IsReplica)) {
@ -142,8 +143,7 @@ auto CoordinatorData::ShowReplicas() const -> std::vector<CoordinatorInstanceSta
}
auto CoordinatorData::SetInstanceToMain(std::string instance_name) -> SetInstanceToMainCoordinatorStatus {
// TODO: (andi) test this
std::lock_guard<utils::RWLock> lock{coord_data_lock_};
auto lock = std::lock_guard{coord_data_lock_};
// Find replica we already registered
auto registered_replica = std::find_if(
@ -185,7 +185,7 @@ auto CoordinatorData::SetInstanceToMain(std::string instance_name) -> SetInstanc
}
auto CoordinatorData::RegisterInstance(CoordinatorClientConfig config) -> RegisterInstanceCoordinatorStatus {
std::lock_guard<utils::RWLock> lock{coord_data_lock_};
auto lock = std::lock_guard{coord_data_lock_};
if (std::ranges::any_of(registered_instances_, [&config](const CoordinatorInstance &instance) {
return instance.InstanceName() == config.instance_name;
})) {

7
src/coordination/include/coordination/coordinator_instance.hpp
View File

@ -36,8 +36,7 @@ class CoordinatorInstance {
~CoordinatorInstance() = default;
auto UpdateInstanceStatus() -> bool {
is_alive_ = std::chrono::duration_cast<std::chrono::seconds>(std::chrono::system_clock::now() -
last_response_time_.load(std::memory_order_acquire))
is_alive_ = std::chrono::duration_cast<std::chrono::seconds>(std::chrono::system_clock::now() - last_response_time_)
.count() < CoordinatorClusterConfig::alive_response_time_difference_sec_;
return is_alive_;
}
@ -66,8 +65,8 @@ class CoordinatorInstance {
CoordinatorClient client_;
replication_coordination_glue::ReplicationRole replication_role_;
std::atomic<std::chrono::system_clock::time_point> last_response_time_{};
std::atomic<bool> is_alive_{false};
std::chrono::system_clock::time_point last_response_time_{};
bool is_alive_{false};
friend bool operator==(CoordinatorInstance const &first, CoordinatorInstance const &second) {
return first.client_ == second.client_ && first.replication_role_ == second.replication_role_;