Random EOF errors on HTTP POST requests #1539
Description
Describe the bug
I'm using Docker-based Gatus for monitoring of a few endpoints (POST request with a hardcoded payload) of my service (at a step of 1 minute). From time to time Gatus shows red status and error reason is "EOF", request time is +/- 50ms. To investigate, I created a load test in Go that does the same request using vegeta and it went at tens of thousands of requests per second without a single error from the same node where Gatus works. Then I deployed Uptime Kuma again (from which a year ago I migrated to Gatus) and now for already about month there was not a single failure with Uptime Kuma (but Gatus "fails" every 5-10 minutes) and the same endpoints/payload. So I conclude the bug is in Gatus, unfortunately. I already tried various configuration push-ups (increase concurrency, timeouts etc.) to no avail.
How to fix or debug this problem? I tried enabling debug/verbose logs but there's the same EOF error in the logs.
What do you see?
No response
What do you expect to see?
Gatus does not fail to make a simple POST request with payload.
List the steps that must be taken to reproduce this issue
config (latest iteration, but it was more bare-bones and also failed):
skip-invalid-config-update: true
metrics: true
concurrency: 5
storage:
type: sqlite
path: /data/data.db
connectivity:
checker:
target: 1.1.1.1:53
interval: 30s
ui:
default-sort-by: group
web:
read-buffer-size: 32768
server-api: &server-snippet
method: POST
interval: "60s"
client:
insecure: true
timeout: 15s
body: |
qwerty12345
headers:
Host: "example.com"
X-Api-Key: "${API_KEY}"
conditions:
- "[STATUS] == 200"
- "[BODY].success == true"
endpoints:
- name: "Server asdf API"
url: "https://123.456.789.012/my-endpoint"
<<: *server-snippet
alerts:
- type: pushover
description: "Server API failed"
Version
stable
Additional information
No response