Hello team, I have some services and kong is reporting 502 errors in POST request, it is a random issue, the services are working normally and randomly appears a 502… Do you know why is that? Kong version 2.1.3 regards

Hi, 502 is usually the result of a bad response from an upstream server. Do you know which Kong service it is associated with? You can try to add the header “Kong-Debug: 1” to see which route/service it is hitting and you can then try to hit that upstream service directly to see if it can be reproduced. Other than that, you can look to the Kong logs for any additional detail. If this is a test environment you may get some more detail out of the log_level setting to debug. Any other errors you see may help if you can post them.

Hello Joedas, Thank you so much for your response. Regarding your questions, Yes, I know the service associated to kong, and it’s quite strange, so for instance, it receives 1.5k request responding with a 200 and randomly during these 1.5k give 20 502… What I don’t know if the issue can come from the service itself or in some part of the communication between Kong and the services it’s broken… I set the header, and yes, the info provided is ok, but as said, it’s quite random and trying to force the error with no success… Also in the logs, I found that message: Connection reset by peer Do you know if that error if from the service or in kong? FYI: Currently Kong is running on kubernetes. Cheers

more info: We have more than one upstream in the end, we are using services and routes (kong 2.1.3) and this is happening in almost all of them. we have a a front ahead of the services and there we cannot see the 502 error, we detected in the kong side.

Looks like we have the same issue: [image] 502 & Keepalive connection closed, AWS/CloudFront + classic LB + Kong (K8s) How can I disable keepalive functionality in Kong? It’s not clear to me from the description whether this will solve your problem, but to your specific question, here’s how to disable the upstream keepalive — note however that this is likely to have a performance impact: https://docs.konghq.com/2.1.x/configuration/#upstream_keepalive_pool_size Depending on your upstream’s behavior, you may want to tune the other related options upstream_keepalive_max_requests and upstream_keepalive_idle_ti… Did not found solution for it. Disabling keepalive feature on Kong did not help.

Hello, We have been sniffing traffic in the kong pods and we cannot see any 502, we only can see error 500 and we have detect some 502 checking the time and comparing it in the data captured… Do you have any idea why I cannot see any 502? Also, we have reduced the values to the next ones: keepalive_pool_size: 128 keepalive_max_requests: 128 keepalive_timeout: 5 Regards

[image] 502 & Keepalive connection closed, AWS/CloudFront + classic LB + Kong (K8s) An HTTP 502 error means that a server received an invalid response from another server. In this case, Kong received an invalid response from your upstream. It is all explained in the log really: upstream prematurely closed connection while reading response headers. Your upstream prematurely closed the connection while Kong was patiently expecting its response. In order to resolve this, you will have to investigate this from the upstream, see if any errors occurred there (since the issue arises …

Random 502 in POST request

General

thibaultcha November 24, 2020, 7:20pm 7

Topic		Replies	Views
502 & Keepalive connection closed, AWS/CloudFront + classic LB + Kong (K8s) General	6	4484	November 24, 2020
104: Connection reset by peer while reading response header from upstream Questions	7	45862	March 19, 2018
Getting 502 as a response from kong, the request has not reached my upstream server Questions	0	94	October 7, 2024
Kong 1.4.0 - Getting random 503 status codes on requests Questions kubernetes	9	5938	July 18, 2020
Need help with resolving intermittent 502 errors Questions	1	377	June 18, 2024

Random 502 in POST request

Related topics