Skip to content
体验新版
项目
组织
正在加载...
登录
切换导航
打开侧边栏
milvus
milvus
提交
20936f78
M
milvus
项目概览
milvus
/
milvus
大约 1 年 前同步成功
通知
261
Star
22476
Fork
2472
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
0
列表
看板
标记
里程碑
合并请求
0
DevOps
流水线
流水线任务
计划
Wiki
0
Wiki
分析
仓库
DevOps
项目成员
Pages
M
milvus
项目概览
项目概览
详情
发布
仓库
仓库
文件
提交
分支
标签
贡献者
分支图
比较
Issue
0
Issue
0
列表
看板
标记
里程碑
合并请求
0
合并请求
0
Pages
DevOps
DevOps
流水线
流水线任务
计划
分析
分析
仓库分析
DevOps
Wiki
0
Wiki
成员
成员
收起侧边栏
关闭侧边栏
动态
分支图
创建新Issue
流水线任务
提交
Issue看板
体验新版 GitCode,发现更多精彩内容 >>
未验证
提交
20936f78
编写于
5月 05, 2022
作者:
Y
yah01
提交者:
GitHub
5月 05, 2022
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
Fix LoadBalance failed to sync segments to shard leader when the leader is offline (#16770)
Signed-off-by:
N
yah01
<
yang.cen@zilliz.com
>
上级
333b439a
变更
3
隐藏空白更改
内联
并排
Showing
3 changed file
with
21 addition
and
7 deletion
+21
-7
internal/querycoord/query_coord.go
internal/querycoord/query_coord.go
+1
-0
internal/querycoord/task.go
internal/querycoord/task.go
+12
-7
internal/querycoord/task_scheduler.go
internal/querycoord/task_scheduler.go
+8
-0
未找到文件。
internal/querycoord/query_coord.go
浏览文件 @
20936f78
...
@@ -474,6 +474,7 @@ func (qc *QueryCoord) handleNodeEvent(ctx context.Context) {
...
@@ -474,6 +474,7 @@ func (qc *QueryCoord) handleNodeEvent(ctx context.Context) {
log
.
Error
(
"unable to allcoate node"
,
zap
.
Int64
(
"nodeID"
,
serverID
),
zap
.
Error
(
err
))
log
.
Error
(
"unable to allcoate node"
,
zap
.
Int64
(
"nodeID"
,
serverID
),
zap
.
Error
(
err
))
}
}
qc
.
metricsCacheManager
.
InvalidateSystemInfoMetrics
()
qc
.
metricsCacheManager
.
InvalidateSystemInfoMetrics
()
case
sessionutil
.
SessionDelEvent
:
case
sessionutil
.
SessionDelEvent
:
serverID
:=
event
.
Session
.
ServerID
serverID
:=
event
.
Session
.
ServerID
log
.
Info
(
"get a del event after QueryNode down"
,
zap
.
Int64
(
"nodeID"
,
serverID
))
log
.
Info
(
"get a del event after QueryNode down"
,
zap
.
Int64
(
"nodeID"
,
serverID
))
...
...
internal/querycoord/task.go
浏览文件 @
20936f78
...
@@ -78,6 +78,7 @@ type task interface {
...
@@ -78,6 +78,7 @@ type task interface {
preExecute
(
ctx
context
.
Context
)
error
preExecute
(
ctx
context
.
Context
)
error
execute
(
ctx
context
.
Context
)
error
execute
(
ctx
context
.
Context
)
error
postExecute
(
ctx
context
.
Context
)
error
postExecute
(
ctx
context
.
Context
)
error
globalPostExecute
(
ctx
context
.
Context
)
error
// execute after all child task completed
reschedule
(
ctx
context
.
Context
)
([]
task
,
error
)
reschedule
(
ctx
context
.
Context
)
([]
task
,
error
)
rollBack
(
ctx
context
.
Context
)
[]
task
rollBack
(
ctx
context
.
Context
)
[]
task
waitToFinish
()
error
waitToFinish
()
error
...
@@ -273,6 +274,10 @@ func (bt *baseTask) getResultInfo() *commonpb.Status {
...
@@ -273,6 +274,10 @@ func (bt *baseTask) getResultInfo() *commonpb.Status {
return
proto
.
Clone
(
bt
.
result
)
.
(
*
commonpb
.
Status
)
return
proto
.
Clone
(
bt
.
result
)
.
(
*
commonpb
.
Status
)
}
}
func
(
bt
*
baseTask
)
globalPostExecute
(
ctx
context
.
Context
)
error
{
return
nil
}
func
(
bt
*
baseTask
)
updateTaskProcess
()
{
func
(
bt
*
baseTask
)
updateTaskProcess
()
{
// TODO::
// TODO::
}
}
...
@@ -2222,13 +2227,6 @@ func (lbt *loadBalanceTask) getReplica(nodeID, collectionID int64) (*milvuspb.Re
...
@@ -2222,13 +2227,6 @@ func (lbt *loadBalanceTask) getReplica(nodeID, collectionID int64) (*milvuspb.Re
}
}
func
(
lbt
*
loadBalanceTask
)
postExecute
(
context
.
Context
)
error
{
func
(
lbt
*
loadBalanceTask
)
postExecute
(
context
.
Context
)
error
{
err
:=
syncReplicaSegments
(
lbt
.
ctx
,
lbt
.
cluster
,
lbt
.
getChildTask
())
if
err
!=
nil
{
log
.
Error
(
"loadBalanceTask: failed to sync replica segments to shard leaders"
,
zap
.
Int64
(
"taskID"
,
lbt
.
getTaskID
()),
zap
.
Error
(
err
))
}
if
lbt
.
getResultInfo
()
.
ErrorCode
!=
commonpb
.
ErrorCode_Success
{
if
lbt
.
getResultInfo
()
.
ErrorCode
!=
commonpb
.
ErrorCode_Success
{
lbt
.
clearChildTasks
()
lbt
.
clearChildTasks
()
}
}
...
@@ -2299,6 +2297,13 @@ func (lbt *loadBalanceTask) postExecute(context.Context) error {
...
@@ -2299,6 +2297,13 @@ func (lbt *loadBalanceTask) postExecute(context.Context) error {
return
nil
return
nil
}
}
func
(
lbt
*
loadBalanceTask
)
globalPostExecute
(
ctx
context
.
Context
)
error
{
if
len
(
lbt
.
getChildTask
())
>
0
{
return
syncReplicaSegments
(
ctx
,
lbt
.
cluster
,
lbt
.
getChildTask
())
}
return
nil
}
func
assignInternalTask
(
ctx
context
.
Context
,
func
assignInternalTask
(
ctx
context
.
Context
,
parentTask
task
,
meta
Meta
,
cluster
Cluster
,
parentTask
task
,
meta
Meta
,
cluster
Cluster
,
loadSegmentRequests
[]
*
querypb
.
LoadSegmentsRequest
,
loadSegmentRequests
[]
*
querypb
.
LoadSegmentsRequest
,
...
...
internal/querycoord/task_scheduler.go
浏览文件 @
20936f78
...
@@ -684,6 +684,14 @@ func (scheduler *TaskScheduler) scheduleLoop() {
...
@@ -684,6 +684,14 @@ func (scheduler *TaskScheduler) scheduleLoop() {
}
}
}
}
err
=
triggerTask
.
globalPostExecute
(
triggerTask
.
traceCtx
())
if
err
!=
nil
{
log
.
Error
(
"scheduleLoop: failed to execute globalPostExecute() of task"
,
zap
.
Int64
(
"taskID"
,
triggerTask
.
getTaskID
()),
zap
.
Error
(
err
))
triggerTask
.
setResultInfo
(
err
)
}
err
=
removeTaskFromKVFn
(
triggerTask
)
err
=
removeTaskFromKVFn
(
triggerTask
)
if
err
!=
nil
{
if
err
!=
nil
{
log
.
Error
(
"scheduleLoop: error when remove trigger and internal tasks from etcd"
,
zap
.
Int64
(
"triggerTaskID"
,
triggerTask
.
getTaskID
()),
zap
.
Error
(
err
))
log
.
Error
(
"scheduleLoop: error when remove trigger and internal tasks from etcd"
,
zap
.
Int64
(
"triggerTaskID"
,
triggerTask
.
getTaskID
()),
zap
.
Error
(
err
))
...
...
编辑
预览
Markdown
is supported
0%
请重试
或
添加新附件
.
添加附件
取消
You are about to add
0
people
to the discussion. Proceed with caution.
先完成此消息的编辑!
取消
想要评论请
注册
或
登录