tokers commented on issue #3616: URL: https://github.com/apache/apisix/issues/3616#issuecomment-785789041
> > Hi, @yocaning > > You need to pay attention to your configuration. The `http_path` of the active health check should be consistent with the uri that provides the service. Secondly, the value of `"timeout":1` is recommended to be set larger. > > This is an example, you can refer to it > > 1、Add a route and enable health check > > The 1981 and 1982 ports are healthy, but the 1985 port is unhealthy. > > ```shell > > curl http://127.0.0.1:9080/apisix/admin/routes/1 -H 'X-API-KEY: edd1c9f034335f136f87ad84b625c8f1' -X PUT -d ' > > { > > "id":"1", > > "update_time":1613734165, > > "upstream":{ > > "checks":{ > > "active":{ > > "healthy":{ > > "interval":3, > > "successes":2 > > }, > > "unhealthy":{ > > "interval":10, > > "http_failures":2 > > }, > > "timeout":10, > > "http_path":"/hello", > > "type":"http", > > "req_headers":[ > > "Content-Seq:3" > > ] > > }, > > "passive":{ > > "timeout":5, > > "healthy":{ > > "http_statuses":[ > > 200, > > 201 > > ], > > "successes":3 > > }, > > "type":"http", > > "unhealthy":{ > > "http_statuses":[ > > 500, > > 504 > > ], > > "timeouts":3, > > "tcp_failures":1, > > "http_failures":1 > > } > > } > > }, > > "nodes":{ > > "127.0.0.1:1980":1, > > "127.0.0.1:1981":1, > > "127.0.0.1:1985":1 > > }, > > "timeout":{ > > "connect":3, > > "read":3, > > "send":3 > > }, > > "pass_host":"pass", > > "hash_on":"vars", > > "type":"roundrobin" > > }, > > "methods":[ > > "PUT", > > "GET" > > ], > > "status":1, > > "create_time":1613722689, > > "priority":0, > > "uri":"/hello" > > }' > > ``` > > > > > > 2、Access request and view error.log > > ``` > > $ curl http://127.0.0.1:9080/hello > > hello 1980 > > > > $ tail -f error.log > > 2021/02/22 15:41:42 [warn] 23035#23035: *66190 [lua] healthcheck.lua:1107: log(): [healthcheck] (upstream#/apisix/routes/1) unhealthy TCP increment (1/2) for '(127.0.0.1:1985)', context: ngx.timer, client: 127.0.0.1, server: 0.0.0.0:9080 > > 2021/02/22 15:41:48 [warn] 23035#23035: *66445 [lua] healthcheck.lua:1107: log(): [healthcheck] (upstream#/apisix/routes/1) unhealthy TCP increment (2/2) for '(127.0.0.1:1985)', context: ngx.timer, client: 127.0.0.1, server: 0.0.0.0:9080 > > > > 2021/02/22 15:42:56 [warn] 23036#23036: *69367 [lua] healthcheck.lua:1107: log(): [healthcheck] (upstream#/apisix/routes/1) healthy SUCCESS increment (1/2) for '(127.0.0.1:1980)', context: ngx.timer, client: 127.0.0.1, server: 0.0.0.0:9080 > > 2021/02/22 15:42:56 [warn] 23036#23036: *69367 [lua] healthcheck.lua:1107: log(): [healthcheck] (upstream#/apisix/routes/1) healthy SUCCESS increment (1/2) for '(127.0.0.1:1981)', context: ngx.timer, client: 127.0.0.1, server: 0.0.0.0:9080 > > ``` > > > > > > 3、When the port of `1985` recovered health and view error.log > > When an unhealthy node is restored to a healthy node, it will be automatically marked as healthy. > > ``` > > 2021/02/22 15:57:32 [warn] 23035#23035: *107236 [lua] healthcheck.lua:1107: log(): [healthcheck] (upstream#/apisix/routes/1) healthy SUCCESS increment (2/2) for '(127.0.0.1:1985)', context: ngx.timer, client: 127.0.0.1, server: 0.0.0.0:9080 > > ``` > > What are the requirements for this health check method? when my node is unhealthy , the auto healthy check method work。and response http header status =200 。but still not marked as healthy The active health checker needs more times try to judge whether a node is healthy or unhealthy, not just one time. See https://github.com/apache/apisix/blob/master/apisix/schema_def.lua#L134. You may tweak these args to see the change. ---------------------------------------------------------------- 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: us...@infra.apache.org
