Hello everyone, hope somene here can help.
We have two kong nodes, they are using the same postgres database, so as I understand they forma kong cluster, correct ?
We are also using upstreams for our services, but only with active healthchecks for now. The configuration looks like this:
{
"hash_on": "none",
"healthchecks": {
"passive": {
"healthy": {
"http_statuses": [
200,
201,
202,
203,
204,
205,
206,
207,
208,
226,
300,
301,
302,
303,
304,
305,
306,
307,
308
],
"successes": 0
},
"unhealthy": {
"tcp_failures": 0,
"timeouts": 0,
"http_failures": 0,
"http_statuses": [
429,
500,
503
]
}
},
"active": {
"timeout": 5,
"healthy": {
"interval": 120,
"successes": 3,
"http_statuses": [
200
]
},
"http_path": "/getversion/",
"concurrency": 10,
"unhealthy": {
"timeouts": 3,
"http_statuses": [
404,
429,
500,
501,
502,
503,
504,
505
],
"http_failures": 3,
"tcp_failures": 3,
"interval": 10
}
}
},
"slots": 10000,
"id": "183c79dc-1ff5-4ef9-895f-8c3a73a48ecb",
"hash_on_cookie_path": "/",
"created_at": 1544470425359,
"name": "API",
"hash_fallback": "none"
}
The configuration is exactly the same between the nodes, as it should be, since they are using the same database, right ?
When I request for the health of the upstream I get all targets as “HEALTHY” in one node but in the other one I get “HEALTHCHECKS_OFF” for all targets.
When I execute a request for this service through the node that has “HEALTHCHECKS_OFF” it works just fine.
Is there something we’re doing wrong ? I thought that all targets should be marked as “HEALTHY” in all nodes. We need this to configure our alerts correctly.