401 lines
15 KiB
C++
401 lines
15 KiB
C++
/**
|
|
* Copyright (c) 2021 OceanBase
|
|
* OceanBase CE is licensed under Mulan PubL v2.
|
|
* You can use this software according to the terms and conditions of the Mulan PubL v2.
|
|
* You may obtain a copy of Mulan PubL v2 at:
|
|
* http://license.coscl.org.cn/MulanPubL-2.0
|
|
* THIS SOFTWARE IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND,
|
|
* EITHER EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO NON-INFRINGEMENT,
|
|
* MERCHANTABILITY OR FIT FOR A PARTICULAR PURPOSE.
|
|
* See the Mulan PubL v2 for more details.
|
|
*/
|
|
|
|
#define USING_LOG_PREFIX RS
|
|
|
|
#include "ob_partition_table_util.h"
|
|
#include "ob_root_utils.h"
|
|
|
|
#include "lib/container/ob_se_array_iterator.h"
|
|
#include "share/partition_table/ob_partition_info.h"
|
|
#include "share/partition_table/ob_partition_table_iterator.h"
|
|
#include "share/partition_table/ob_partition_table_operator.h"
|
|
#include "share/schema/ob_schema_getter_guard.h"
|
|
#include "share/schema/ob_multi_version_schema_service.h"
|
|
#include "share/ob_global_stat_proxy.h"
|
|
#include "ob_server_manager.h"
|
|
#include "ob_zone_manager.h"
|
|
#include "observer/ob_server_struct.h"
|
|
|
|
namespace oceanbase {
|
|
|
|
using namespace common;
|
|
using namespace share;
|
|
using namespace share::schema;
|
|
|
|
namespace rootserver {
|
|
|
|
template <typename T>
|
|
T first_param_percnetage(const T first, const T second)
|
|
{
|
|
T percentage = 100;
|
|
const T sum = first + second;
|
|
if (0 != sum) {
|
|
percentage = first * 100 / sum;
|
|
}
|
|
return percentage;
|
|
}
|
|
|
|
int64_t ObPartitionTableUtil::MergeProgress::get_merged_partition_percentage() const
|
|
{
|
|
return first_param_percnetage(merged_partition_cnt_, unmerged_partition_cnt_);
|
|
}
|
|
|
|
int64_t ObPartitionTableUtil::MergeProgress::get_merged_data_percentage() const
|
|
{
|
|
return first_param_percnetage(merged_data_size_, unmerged_data_size_);
|
|
}
|
|
|
|
ObPartitionTableUtil::ObPartitionTableUtil()
|
|
: inited_(false), zone_mgr_(NULL), schema_service_(NULL), pt_(NULL), server_mgr_(NULL), sql_proxy_(NULL)
|
|
{}
|
|
|
|
ObPartitionTableUtil::~ObPartitionTableUtil()
|
|
{}
|
|
|
|
int ObPartitionTableUtil::init(ObZoneManager& zone_mgr, ObMultiVersionSchemaService& schema_service,
|
|
ObPartitionTableOperator& pt, ObServerManager& server_mgr, common::ObMySQLProxy& sql_proxy)
|
|
{
|
|
int ret = OB_SUCCESS;
|
|
if (inited_) {
|
|
ret = OB_INIT_TWICE;
|
|
LOG_WARN("init twice", K(ret));
|
|
} else {
|
|
zone_mgr_ = &zone_mgr;
|
|
schema_service_ = &schema_service;
|
|
pt_ = &pt;
|
|
server_mgr_ = &server_mgr;
|
|
sql_proxy_ = &sql_proxy;
|
|
inited_ = true;
|
|
}
|
|
return ret;
|
|
}
|
|
|
|
int ObPartitionTableUtil::check_merge_progress(
|
|
const volatile bool& stop, const int64_t version, ObZoneMergeProgress& all_progress, bool& all_majority_merged)
|
|
{
|
|
int ret = OB_SUCCESS;
|
|
const int64_t start_time = ObTimeUtility::current_time();
|
|
ObSEArray<ObZone, MAX_ZONE_NUM> all_zones;
|
|
int64_t standby_snapshot_frozen_version = 0;
|
|
int64_t standby_snapshot_schema_version = 0;
|
|
ObClusterType cluster_type;
|
|
if (!inited_) {
|
|
ret = OB_NOT_INIT;
|
|
LOG_WARN("partition table util not inited", K(ret));
|
|
} else if (stop) {
|
|
ret = OB_CANCELED;
|
|
} else if (version <= 0 || all_progress.error()) {
|
|
ret = OB_INVALID_ARGUMENT;
|
|
LOG_WARN("invalid argument", K(ret), K(version), "array_error", all_progress.error());
|
|
} else if (FALSE_IT(cluster_type = ObClusterInfoGetter::get_cluster_type_v2())) {
|
|
// nothing todo
|
|
} else {
|
|
all_majority_merged = true;
|
|
all_progress.reset();
|
|
if (OB_FAIL(zone_mgr_->get_zone(all_zones))) {
|
|
LOG_WARN("get zone failed", K(ret));
|
|
} else {
|
|
// init %all_progress
|
|
MergeProgress zone_progress;
|
|
for (int64_t i = 0; OB_SUCC(ret) && i < all_zones.count(); ++i) {
|
|
zone_progress.zone_ = all_zones.at(i);
|
|
if (OB_FAIL(all_progress.push_back(zone_progress))) {
|
|
LOG_WARN("add zone merge progress to array failed", K(ret));
|
|
}
|
|
}
|
|
std::sort(all_progress.begin(), all_progress.end());
|
|
}
|
|
}
|
|
bool need_check_root_table_integrated = true;
|
|
if (OB_SUCC(ret)) {
|
|
ObGlobalStatProxy global_proxy(*sql_proxy_);
|
|
if (STANDBY_CLUSTER != cluster_type) {
|
|
// nothing todo
|
|
} else if (OB_FAIL(global_proxy.get_schema_snapshot_version(
|
|
standby_snapshot_schema_version, standby_snapshot_frozen_version))) {
|
|
LOG_WARN("fail to get schema snapshot versoin", KR(ret));
|
|
} else if (standby_snapshot_frozen_version == 0 && version <= 1) {
|
|
need_check_root_table_integrated = false;
|
|
} else if (version <= standby_snapshot_frozen_version) {
|
|
need_check_root_table_integrated = false;
|
|
}
|
|
}
|
|
ObSchemaGetterGuard schema_guard;
|
|
ObPartitionTableIterator iter;
|
|
bool ignore_row_checksum = true;
|
|
if (OB_FAIL(ret)) {
|
|
} else if (OB_FAIL(iter.init(*pt_, *schema_service_, ignore_row_checksum))) {
|
|
LOG_WARN("partition table iterator init failed", K(ret));
|
|
} else if (OB_FAIL(iter.get_filters().set_only_alive_server(*server_mgr_))) {
|
|
LOG_WARN("set filter failed", K(ret));
|
|
} else if (OB_FAIL(iter.get_filters().set_replica_status(REPLICA_STATUS_NORMAL))) {
|
|
LOG_WARN("set filter failed", K(ret));
|
|
} else if (OB_FAIL(iter.get_filters().filter_rebuild_replica())) {
|
|
LOG_WARN("fail to set filter", K(ret));
|
|
} else if (OB_FAIL(iter.get_filters().filter_restore_replica())) {
|
|
LOG_WARN("fail to set filter", K(ret));
|
|
} else if (OB_FAIL(schema_service_->get_schema_guard(schema_guard))) {
|
|
LOG_WARN("failed to get schema guard", K(ret));
|
|
} else {
|
|
int64_t majority = 0;
|
|
int64_t full_replica_num = OB_INVALID_COUNT;
|
|
int64_t all_replica_num = OB_INVALID_COUNT;
|
|
int64_t paxos_replica_num = OB_INVALID_COUNT;
|
|
uint64_t last_partition_entity_id = OB_INVALID_ID;
|
|
ObPartitionInfo partition;
|
|
bool need_check = true;
|
|
while (!stop && OB_SUCC(ret) && OB_SUCC(iter.next(partition))) {
|
|
if (OB_INVALID_ID == last_partition_entity_id || last_partition_entity_id != partition.get_table_id()) {
|
|
const uint64_t this_partition_entity_id = partition.get_table_id();
|
|
bool is_restore = false;
|
|
const uint64_t tenant_id = partition.get_tenant_id();
|
|
if (OB_FAIL(schema_guard.check_tenant_is_restore(tenant_id, is_restore))) {
|
|
LOG_WARN("fail to check tenant is restore", K(ret), K(tenant_id));
|
|
} else if (is_restore && !is_inner_table(this_partition_entity_id)) {
|
|
need_check = false; // user table do not suspend merge during restore
|
|
} else if (!is_tablegroup_id(this_partition_entity_id)) {
|
|
const share::schema::ObTableSchema* table_schema = nullptr;
|
|
if (OB_FAIL(schema_guard.get_table_schema(this_partition_entity_id, table_schema))) {
|
|
LOG_WARN("fail to get table schema", K(ret), "table_id", this_partition_entity_id);
|
|
} else if (nullptr == table_schema) {
|
|
need_check = false; // ignore the table does not exist
|
|
} else if (OB_FAIL(get_associated_replica_num(schema_guard,
|
|
*table_schema,
|
|
paxos_replica_num,
|
|
full_replica_num,
|
|
all_replica_num,
|
|
majority))) {
|
|
LOG_WARN("fail to get associated replica num", K(ret));
|
|
} else {
|
|
last_partition_entity_id = this_partition_entity_id;
|
|
need_check = !(table_schema->is_index_table() && table_schema->is_global_index_table() &&
|
|
!table_schema->can_read_index());
|
|
}
|
|
} else {
|
|
const share::schema::ObTablegroupSchema* tg_schema = nullptr;
|
|
if (OB_FAIL(schema_guard.get_tablegroup_schema(this_partition_entity_id, tg_schema))) {
|
|
LOG_WARN("fail to get tablegroup schema", K(ret), "tg_id", this_partition_entity_id);
|
|
} else if (nullptr == tg_schema) {
|
|
need_check = false; // ignore the tablegroup does not exist
|
|
} else if (OB_FAIL(get_associated_replica_num(
|
|
schema_guard, *tg_schema, paxos_replica_num, full_replica_num, all_replica_num, majority))) {
|
|
LOG_WARN("fail to get associated replica num", K(ret));
|
|
} else {
|
|
last_partition_entity_id = this_partition_entity_id;
|
|
need_check = true;
|
|
}
|
|
}
|
|
}
|
|
if (OB_SUCC(ret)) {
|
|
// check the integrity for every replica
|
|
// at lease one full replica is required, number of paxos replicas shall no more than majority
|
|
int64_t full_cnt = 0;
|
|
int64_t paxos_cnt = 0;
|
|
FOREACH_CNT_X(r, partition.get_replicas_v2(), OB_SUCC(ret))
|
|
{
|
|
if (OB_ISNULL(r)) {
|
|
ret = OB_ERR_UNEXPECTED;
|
|
LOG_WARN("get invalid replica", K(ret), K(r));
|
|
} else if (!r->is_in_service()) {
|
|
// noting todo
|
|
} else if (ObReplicaTypeCheck::is_paxos_replica_V2(r->replica_type_)) {
|
|
paxos_cnt++;
|
|
if (REPLICA_TYPE_FULL == r->replica_type_) {
|
|
full_cnt++;
|
|
}
|
|
}
|
|
} // end FOREACH_CNT_X(
|
|
if (OB_SUCC(ret) && need_check_root_table_integrated && need_check) {
|
|
// ignore the replica integrity check for unavailable index
|
|
// since replicas in standby cluster are not copied until index is available
|
|
if (full_cnt <= 0 || paxos_cnt < majority) {
|
|
const bool is_cluster_private_table =
|
|
!GCTX.is_standby_cluster() || ObMultiClusterUtil::is_cluster_private_table(partition.get_table_id());
|
|
ret = OB_ROOT_NOT_INTEGRATED;
|
|
if (is_cluster_private_table) {
|
|
LOG_ERROR("partition does not have enough alive normal replicas",
|
|
K(majority),
|
|
K(partition),
|
|
K(paxos_cnt),
|
|
K(paxos_cnt));
|
|
} else {
|
|
LOG_WARN("partition does not have enough alive normal replicas",
|
|
K(majority),
|
|
K(partition),
|
|
K(paxos_cnt),
|
|
K(paxos_cnt));
|
|
}
|
|
}
|
|
}
|
|
} // end if
|
|
if (OB_FAIL(ret)) {
|
|
} else if (!need_check) {
|
|
// no need to check when schema not exist, or unreadable global index
|
|
} else {
|
|
// check all replica merged to %version
|
|
int64_t merged_cnt = 0;
|
|
int64_t full_merged_cnt = 0;
|
|
FOREACH_CNT_X(r, partition.get_replicas_v2(), OB_SUCCESS == ret)
|
|
{
|
|
ObZoneMergeProgress::iterator p = std::lower_bound(all_progress.begin(), all_progress.end(), r->zone_);
|
|
if (p != all_progress.end() && p->zone_ == r->zone_) {
|
|
if (REPLICA_TYPE_LOGONLY == r->replica_type_) {
|
|
// nothing todo
|
|
} else {
|
|
if (p->smallest_data_version_ <= 0 || p->smallest_data_version_ > r->data_version_) {
|
|
p->smallest_data_version_ = r->data_version_;
|
|
// only log the first old version replica
|
|
if (r->data_version_ < version - 1) {
|
|
LOG_WARN("replica data version is too old", K(version), "replica", *r);
|
|
}
|
|
}
|
|
if (r->data_version_ < version) {
|
|
// only log the first replica not merged
|
|
if (p->unmerged_partition_cnt_ == 0) {
|
|
LOG_INFO("replica not merged to version", K(version), "replica", *r);
|
|
}
|
|
p->unmerged_partition_cnt_++;
|
|
p->unmerged_data_size_ += r->data_size_;
|
|
} else {
|
|
if (REPLICA_TYPE_FULL == r->replica_type_) {
|
|
full_merged_cnt++;
|
|
}
|
|
p->merged_partition_cnt_++;
|
|
p->merged_data_size_ += r->data_size_;
|
|
merged_cnt++;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
if (OB_SUCC(ret)) {
|
|
// standard for most merged:
|
|
// 1 the majority of full replicas merge finished
|
|
// 2 the majority of f+r replicas merge finished, ignore readonly@all locality
|
|
int64_t replica_count_with_sstable = all_replica_num - (paxos_replica_num - full_replica_num);
|
|
if (all_majority_merged && (full_merged_cnt < rootserver::majority(full_replica_num) ||
|
|
merged_cnt < rootserver::majority(replica_count_with_sstable))) {
|
|
all_majority_merged = false;
|
|
// only log the first partition which majority replicas not merged
|
|
LOG_INFO("partition's majority replicas not merged", K(version), K(majority), K(merged_cnt), K(partition));
|
|
}
|
|
}
|
|
}
|
|
}
|
|
if (OB_ITER_END == ret) {
|
|
ret = OB_SUCCESS;
|
|
}
|
|
if (stop) {
|
|
ret = OB_CANCELED;
|
|
LOG_WARN("stop", K(ret));
|
|
}
|
|
}
|
|
const int64_t end_time = ObTimeUtility::current_time();
|
|
if (OB_SUCC(ret)) {
|
|
LOG_INFO("check merge progress success",
|
|
K(version),
|
|
K(all_majority_merged),
|
|
K(all_progress),
|
|
"checking_time_us",
|
|
end_time - start_time);
|
|
} else {
|
|
LOG_WARN("check merge progress failed", K(ret), K(version));
|
|
}
|
|
|
|
return ret;
|
|
}
|
|
|
|
/*
|
|
int ObPartitionTableUtil::set_leader_backup_flag(
|
|
const volatile bool &stop,
|
|
const bool flag,
|
|
ObLeaderInfoArray *leader_info_array)
|
|
{
|
|
int ret = OB_SUCCESS;
|
|
ObPartitionTableIterator iter;
|
|
bool ignore_row_checksum = true;
|
|
|
|
if (NULL != leader_info_array) {
|
|
leader_info_array->reset();
|
|
}
|
|
if (!inited_) {
|
|
ret = OB_NOT_INIT;
|
|
LOG_WARN("not init");
|
|
} else if (OB_FAIL(iter.init(*pt_, *schema_service_, ignore_row_checksum))) {
|
|
LOG_WARN("partition table iterator init failed", K(ret));
|
|
} else {
|
|
ObPartitionInfo partition;
|
|
while (!stop && OB_SUCCESS == ret) {
|
|
if (OB_FAIL(iter.next(partition))) {
|
|
if (OB_ITER_END == ret) {
|
|
ret = OB_SUCCESS;
|
|
} else {
|
|
LOG_WARN("iterator partition table failed", K(ret));
|
|
}
|
|
break;
|
|
}
|
|
|
|
bool exist_flag = false;
|
|
FOREACH_CNT_X(r, partition.get_replicas_v2(), OB_SUCCESS == ret && !exist_flag) {
|
|
if (r->is_original_leader_) {
|
|
exist_flag = true;
|
|
}
|
|
} //end foreach
|
|
if (OB_SUCC(ret)) {
|
|
const ObPartitionReplica *r = NULL;
|
|
if (OB_FAIL(partition.find_leader_v2(r))) {
|
|
LOG_WARN("fail to find leader", K(ret), K(partition));
|
|
if (OB_ENTRY_NOT_EXIST == ret) {
|
|
ret = OB_SUCCESS;
|
|
}
|
|
} else if (OB_ISNULL(r)) {
|
|
ret = OB_ERR_UNEXPECTED;
|
|
LOG_WARN("get invalid leader", K(ret), K(r), K(partition));
|
|
} else if (NULL != leader_info_array) {
|
|
bool found = false;
|
|
for (int64_t i = 0; OB_SUCC(ret) && i < leader_info_array->count(); ++i) {
|
|
if (leader_info_array->at(i).zone_ == r->zone_) {
|
|
found = true;
|
|
++leader_info_array->at(i).leader_count_;
|
|
break;
|
|
}
|
|
} //end for
|
|
|
|
if (OB_SUCC(ret) && !found) {
|
|
ObZoneLeaderInfo info;
|
|
info.zone_ = r->zone_;
|
|
info.leader_count_ = 1;
|
|
if (OB_FAIL(leader_info_array->push_back(info))) {
|
|
LOG_WARN("failed to add zone leader info", K(ret));
|
|
}
|
|
}
|
|
}
|
|
}
|
|
if (flag != exist_flag){
|
|
if (OB_FAIL(pt_->set_original_leader(
|
|
partition.get_table_id(), partition.get_partition_id(), flag))) {
|
|
LOG_WARN("set leader backup failed", K(ret), K(partition));
|
|
}
|
|
}
|
|
}
|
|
if (stop) {
|
|
ret = OB_CANCELED;
|
|
LOG_WARN("stop", K(ret));
|
|
}
|
|
}
|
|
|
|
return ret;
|
|
}*/
|
|
|
|
} // end namespace rootserver
|
|
} // end namespace oceanbase
|