MockingJayWong opened a new issue #1919:
URL: https://github.com/apache/servicecomb-java-chassis/issues/1919
### 微服务调用超时探究
> 本服务A目前调用服务B同一接口偶现超时情况, 通过定位metric日志和access日志，目前发现两种场景。
* 本服务内报超时,
但是对方的access.log无记录日志。对应的metric日志描述如下，getConnect阶段就失效了。检查了client端连接池的配置，单个ip:port连接数最大限制配置的是50，而且查看对应的连接数情况，未找出连接过多而无法获取连接的异常嫌疑。___这块的原因还有什么可能性？___
```
os:
cpu:
all usage: 0.60% all idle: 99.40% process: 3.55%
net:
send(Bps) recv(Bps) send(pps) recv(pps) interface
5 5 0 0 lo
111.120K 42.174K 268 336 eth0
vertx:
instances:
name eventLoopContext-created
registry 0
transport 0
monitor-center 0
config-center 0
transport:
client.endpoints:
remote connectCount disconnectCount queue
connections send(Bps) receive(Bps)
(summary) 76 48 0 99
364 12.579K
server.endpoints:
listen connectCount disconnectCount rejectByLimit
connections send(Bps) receive(Bps)
0.0.0.0:18080 30 20 0 71
17.956K 561
(summary) 30 20 0 71
17.956K 561
---省略----
details:
rest.200:
rest.400:
rest.490:
服务B.schema.对应接口:
prepare : 0.008/0.008 handlersReq : 0.217/0.217
cFiltersReq: 0.022/0.022 sendReq : 0.000/0.000
getConnect : 0.000/0.000 writeBuf : 0.000/0.000
waitResp : 0.000/0.000 wakeConsumer: 0.001/0.001
cFiltersResp: 0.000/0.000 handlersResp: 2.740/2.740
```
* 本服务报超时, 但是对方服务的metric.log, access.log中表现正常
```
access.log duration显示2ms:
[20200810 14:01:46.707][I][tloop-thread-19][accesslog.42]
-10.4.161.25||"POST 对应接口
HTTP/1.1"||200||2||76||10.11.88.79:18080||10/08/2020:14:01:46
+0800||5f30e2ca44821e32||-||-||-
metrics.log:
对应接口:_ prepare: 0.009/0.009 queue : 0.060/0.060
filtersReq : 0.035/0.035 handlersReq: 0.045/0.045_ execute:
1.453/1.453 handlersResp: 0.011/0.011 filtersResp: 0.030/0.030
sendResp : 0.012/0.012_
```
请教下两种情况下的超时分析思路，谢谢。
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
