This is an automated email from the ASF dual-hosted git repository.
morningman pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-doris.git
The following commit(s) were added to refs/heads/master by this push:
new 6be7a62 [Config] Add ignore config to determine whether to continue
to start be when load tablet from header failed. (#3632)
6be7a62 is described below
commit 6be7a6232fd52cba9c9ac32e054d3ffc76210eb0
Author: yangzhg <780531...@qq.com>
AuthorDate: Wed May 20 09:40:50 2020 +0800
[Config] Add ignore config to determine whether to continue to start be
when load tablet from header failed. (#3632)
Add config ignore_load_tablet_failure to determine whether to continue to
start be when load tablet from header failed.
---
be/src/common/config.h | 23 ++++++++++++----------
be/src/olap/data_dir.cpp | 7 ++++++-
docs/en/administrator-guide/config/be_config.md | 5 +++++
docs/zh-CN/administrator-guide/config/be_config.md | 5 +++++
thirdparty/vars.sh | 2 +-
5 files changed, 30 insertions(+), 12 deletions(-)
diff --git a/be/src/common/config.h b/be/src/common/config.h
index ce6a360..b93b72c 100644
--- a/be/src/common/config.h
+++ b/be/src/common/config.h
@@ -34,7 +34,7 @@ namespace config {
// Note that there should at most one ip match this list.
// this is a list in semicolon-delimited format, in CIDR notation, e.g.
10.10.10.0/24
// If no ip match this rule, will choose one randomly.
- CONF_String(priority_networks, "")
+ CONF_String(priority_networks, "");
////
//// tcmalloc gc parameter
@@ -221,7 +221,7 @@ namespace config {
// 仅仅是建议值，当磁盘空间不足时，trash下的文件保存期可不遵守这个参数
CONF_mInt32(trash_file_expire_time_sec, "259200");
// check row nums for BE/CE and schema change. true is open, false is
closed.
- CONF_mBool(row_nums_check, "true")
+ CONF_mBool(row_nums_check, "true");
//file descriptors cache, by default, cache 32768 descriptors
CONF_Int32(file_descriptor_cache_capacity, "32768");
// minimum file descriptor number
@@ -259,7 +259,7 @@ namespace config {
// if compaction of a tablet failed, this tablet should not be chosen to
// compaction until this interval passes.
- CONF_mInt64(min_compaction_failure_interval_sec, "600") // 10 min
+ CONF_mInt64(min_compaction_failure_interval_sec, "600"); // 10 min
// Too many compaction tasks may run out of memory.
// This config is to limit the max concurrency of running compaction tasks.
// -1 means no limit, and the max concurrency will be:
@@ -345,11 +345,11 @@ namespace config {
CONF_Bool(enable_quadratic_probing, "false");
// for pprof
- CONF_String(pprof_profile_dir, "${DORIS_HOME}/log")
+ CONF_String(pprof_profile_dir, "${DORIS_HOME}/log");
// for partition
// CONF_Bool(enable_partitioned_hash_join, "false")
- CONF_Bool(enable_partitioned_aggregation, "true")
+ CONF_Bool(enable_partitioned_aggregation, "true");
// to forward compatibility, will be removed later
CONF_mBool(enable_token_check, "true");
@@ -478,7 +478,7 @@ namespace config {
// The percent of max used capacity of a data dir
CONF_mInt32(storage_flood_stage_usage_percent, "95"); // 95%
// The min bytes that should be left of a data dir
- CONF_mInt64(storage_flood_stage_left_capacity_bytes, "1073741824") // 1GB
+ CONF_mInt64(storage_flood_stage_left_capacity_bytes, "1073741824"); //
1GB
// number of thread for flushing memtable per store
CONF_Int32(flush_thread_num_per_store, "2");
@@ -491,9 +491,9 @@ namespace config {
CONF_String(default_rowset_type, "ALPHA");
// Maximum size of a single message body in all protocols
- CONF_Int64(brpc_max_body_size, "209715200")
+ CONF_Int64(brpc_max_body_size, "209715200");
// Max unwritten bytes in each socket, if the limit is reached,
Socket.Write fails with EOVERCROWDED
- CONF_Int64(brpc_socket_max_unwritten_bytes, "67108864")
+ CONF_Int64(brpc_socket_max_unwritten_bytes, "67108864");
// max number of txns for every txn_partition_map in txn manager
// this is a self protection to avoid too many txns saving in manager
@@ -503,7 +503,7 @@ namespace config {
// this is a an enhancement for better performance to manage tablet
CONF_Int32(tablet_map_shard_size, "1");
- CONF_String(plugin_path, "${DORIS_HOME}/plugin")
+ CONF_String(plugin_path, "${DORIS_HOME}/plugin");
// txn_map_lock shard size, the value is 2^n, n=0,1,2,3,4
// this is a an enhancement for better performance to manage txn
@@ -511,7 +511,10 @@ namespace config {
// txn_lock shard size, the value is 2^n, n=0,1,2,3,4
// this is a an enhancement for better performance to commit and publish
txn
- CONF_Int32(txn_shard_size, "1024")
+ CONF_Int32(txn_shard_size, "1024");
+
+ // Whether to continue to start be when load tablet from header failed.
+ CONF_Bool(ignore_load_tablet_failure, "false");
} // namespace config
diff --git a/be/src/olap/data_dir.cpp b/be/src/olap/data_dir.cpp
index fbcab3b..3d32302 100644
--- a/be/src/olap/data_dir.cpp
+++ b/be/src/olap/data_dir.cpp
@@ -694,19 +694,24 @@ OLAPStatus DataDir::load() {
// create tablet from tablet meta and add it to tablet mgr
LOG(INFO) << "begin loading tablet from meta";
std::set<int64_t> tablet_ids;
- auto load_tablet_func = [this, &tablet_ids](int64_t tablet_id, int32_t
schema_hash,
+ std::set<int64_t> failed_tablet_ids;
+ auto load_tablet_func = [this, &tablet_ids, &failed_tablet_ids](int64_t
tablet_id, int32_t schema_hash,
const std::string& value) ->
bool {
OLAPStatus status = _tablet_manager->load_tablet_from_meta(this,
tablet_id, schema_hash,
value,
false, false);
if (status != OLAP_SUCCESS) {
LOG(WARNING) << "load tablet from header failed. status:" << status
<< ", tablet=" << tablet_id << "." << schema_hash;
+ failed_tablet_ids.insert(tablet_id);
} else {
tablet_ids.insert(tablet_id);
}
return true;
};
OLAPStatus load_tablet_status = TabletMetaManager::traverse_headers(_meta,
load_tablet_func);
+ if (failed_tablet_ids.size() != 0 && !config::ignore_load_tablet_failure) {
+ LOG(FATAL) << "load tablets from header failed, failed tablets size: "
<< failed_tablet_ids.size();
+ }
if (load_tablet_status != OLAP_SUCCESS) {
LOG(WARNING) << "there is failure when loading tablet headers, path:"
<< _path;
} else {
diff --git a/docs/en/administrator-guide/config/be_config.md
b/docs/en/administrator-guide/config/be_config.md
index 89fe9fb..b037b1d 100644
--- a/docs/en/administrator-guide/config/be_config.md
+++ b/docs/en/administrator-guide/config/be_config.md
@@ -398,3 +398,8 @@ Since this is a brpc configuration, users can also modify
this parameter directl
### webserver_port
### write_buffer_size
+
+### ignore_load_tablet_failure
+* Type: boolean
+* Description: Whether to continue to start be when load tablet from header
failed.
+* Default: false
diff --git a/docs/zh-CN/administrator-guide/config/be_config.md
b/docs/zh-CN/administrator-guide/config/be_config.md
index 3cc73b6..02dbd1e 100644
--- a/docs/zh-CN/administrator-guide/config/be_config.md
+++ b/docs/zh-CN/administrator-guide/config/be_config.md
@@ -396,3 +396,8 @@ under the License.
### `webserver_port`
### `write_buffer_size`
+
+### ignore_load_tablet_failure
+* 类型：布尔
+* 描述：用来决定在有tablet 加在失败的情况下是否忽略错误，继续启动be
+* 默认值： false
\ No newline at end of file
diff --git a/thirdparty/vars.sh b/thirdparty/vars.sh
index 29c2a7b..36e64f5 100644
--- a/thirdparty/vars.sh
+++ b/thirdparty/vars.sh
@@ -262,7 +262,7 @@ CROARINGBITMAP_MD5SUM="29602918e6890ffdeed84cb171857046"
# ORC
ORC_DOWNLOAD="https://github.com/apache/orc/archive/rel/release-1.5.8.tar.gz";
ORC_NAME=orc-1.5.8.tar.gz
-ORC_SOURCE=orc-1.5.8
+ORC_SOURCE=orc-rel-release-1.5.8
ORC_MD5SUM="49eb9ce94060b26d4bc3595b2f1efb4c"
# jemalloc
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@doris.apache.org
For additional commands, e-mail: commits-h...@doris.apache.org
