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 5d6a1a7 [Load] support ignoring eovercrowded when tablet sink (#5156)
5d6a1a7 is described below
commit 5d6a1a72903ed042823a0a4d9887f15f65f9fff1
Author: HuangWei <[email protected]>
AuthorDate: Sat Jan 9 23:40:51 2021 +0800
[Load] support ignoring eovercrowded when tablet sink (#5156)
If adding the ignore_eovercrowded flag, the `PTabletWriterAddBatchRequest`
won't failed on `EOVERCROWDED` to avoid load jobs failed in this error.
It only effects the NodeChannel(the load job), other rpc requests will
still check if overcrowded.
---
be/src/common/config.h | 6 ++++-
be/src/exec/tablet_sink.cpp | 27 ++++++++++++++--------
docs/en/administrator-guide/config/be_config.md | 8 +++++++
docs/zh-CN/administrator-guide/config/be_config.md | 8 +++++++
4 files changed, 39 insertions(+), 10 deletions(-)
diff --git a/be/src/common/config.h b/be/src/common/config.h
index 4454c7f..66e52a3 100644
--- a/be/src/common/config.h
+++ b/be/src/common/config.h
@@ -330,8 +330,12 @@ CONF_mInt64(streaming_load_json_max_mb, "100");
CONF_mInt32(streaming_load_rpc_max_alive_time_sec, "1200");
// the timeout of a rpc to open the tablet writer in remote BE.
// short operation time, can set a short timeout
-CONF_mInt32(tablet_writer_open_rpc_timeout_sec, "60");
+CONF_Int32(tablet_writer_open_rpc_timeout_sec, "60");
+// You can ignore brpc error '[E1011]The server is overcrowded' when writing
data.
+CONF_mBool(tablet_writer_ignore_eovercrowded, "false");
+
// OlapTableSink sender's send interval, should be less than the real response
time of a tablet writer rpc.
+// You may need to lower the speed when the sink receiver bes are too busy.
CONF_mInt32(olap_table_sink_send_interval_ms, "1");
// Fragment thread pool
diff --git a/be/src/exec/tablet_sink.cpp b/be/src/exec/tablet_sink.cpp
index 98f5b35..47305d9 100644
--- a/be/src/exec/tablet_sink.cpp
+++ b/be/src/exec/tablet_sink.cpp
@@ -110,6 +110,9 @@ void NodeChannel::open() {
// This ref is for RPC's reference
_open_closure->ref();
_open_closure->cntl.set_timeout_ms(config::tablet_writer_open_rpc_timeout_sec *
1000);
+ if (config::tablet_writer_ignore_eovercrowded) {
+ _open_closure->cntl.ignore_eovercrowded();
+ }
_stub->tablet_writer_open(&_open_closure->cntl, &request,
&_open_closure->result,
_open_closure);
request.release_id();
@@ -294,6 +297,9 @@ void NodeChannel::cancel() {
closure->ref();
closure->cntl.set_timeout_ms(_rpc_timeout_ms);
+ if (config::tablet_writer_ignore_eovercrowded) {
+ closure->cntl.ignore_eovercrowded();
+ }
_stub->tablet_writer_cancel(&closure->cntl, &request, &closure->result,
closure);
request.release_id();
}
@@ -327,6 +333,9 @@ int NodeChannel::try_send_and_fetch_status() {
_add_batch_closure->reset();
_add_batch_closure->cntl.set_timeout_ms(_rpc_timeout_ms);
+ if (config::tablet_writer_ignore_eovercrowded) {
+ _add_batch_closure->cntl.ignore_eovercrowded();
+ }
if (request.eos()) {
for (auto pid : _parent->_partition_ids) {
@@ -674,13 +683,13 @@ Status OlapTableSink::send(RuntimeState* state, RowBatch*
input_batch) {
}
Status OlapTableSink::close(RuntimeState* state, Status close_status) {
- if (_is_closed) {
- /// The close method may be called twice.
- /// In the open_internal() method of plan_fragment_executor,
close is called once.
- /// If an error occurs in this call, it will be called again in
fragment_mgr.
- /// So here we use a flag to prevent repeated close operations.
- return _close_status;
- }
+ if (_is_closed) {
+ /// The close method may be called twice.
+ /// In the open_internal() method of plan_fragment_executor, close is
called once.
+ /// If an error occurs in this call, it will be called again in
fragment_mgr.
+ /// So here we use a flag to prevent repeated close operations.
+ return _close_status;
+ }
Status status = close_status;
if (status.ok()) {
// only if status is ok can we call this
_profile->total_time_counter().
@@ -758,8 +767,8 @@ Status OlapTableSink::close(RuntimeState* state, Status
close_status) {
Expr::close(_output_expr_ctxs, state);
_output_batch.reset();
- _close_status = status;
- _is_closed = true;
+ _close_status = status;
+ _is_closed = true;
return status;
}
diff --git a/docs/en/administrator-guide/config/be_config.md
b/docs/en/administrator-guide/config/be_config.md
index 73e0476..3437639 100644
--- a/docs/en/administrator-guide/config/be_config.md
+++ b/docs/en/administrator-guide/config/be_config.md
@@ -794,6 +794,14 @@ When writing is too frequent and the disk time is
insufficient, you can configur
### `tablet_writer_open_rpc_timeout_sec`
+### `tablet_writer_ignore_eovercrowded`
+
+* Type: bool
+* Description: Used to ignore brpc error '[E1011]The server is overcrowded'
when writing data.
+* Default value: false
+
+When meet '[E1011]The server is overcrowded' error, you can tune the
configuration `brpc_socket_max_unwritten_bytes`, but it can't be modified at
runtime. Set it to `true` to avoid writing failed temporarily. Notice that, it
only effects `write`, other rpc requests will still check if overcrowded.
+
### `tc_free_memory_rate`
### `tc_max_total_thread_cache_bytes`
diff --git a/docs/zh-CN/administrator-guide/config/be_config.md
b/docs/zh-CN/administrator-guide/config/be_config.md
index 2a00307..7cc4c91 100644
--- a/docs/zh-CN/administrator-guide/config/be_config.md
+++ b/docs/zh-CN/administrator-guide/config/be_config.md
@@ -795,6 +795,14 @@ Stream Load 一般适用于导入几个GB以内的数据，不适合导入过大
### `tablet_writer_open_rpc_timeout_sec`
+### `tablet_writer_ignore_eovercrowded`
+
+* 类型：bool
+* 描述：写入时可忽略brpc的'[E1011]The server is overcrowded'错误。
+* 默认值：false
+
+当遇到'[E1011]The server is
overcrowded'的错误时，可以调整配置项`brpc_socket_max_unwritten_bytes`，但这个配置项不能动态调整。所以可通过设置此项为`true`来临时避免写失败。注意，此配置项只影响写流程，其他的rpc请求依旧会检查是否overcrowded。
+
### `tc_free_memory_rate`
### `tc_max_total_thread_cache_bytes`
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
