Http error after update

ESPHome
1

I have 5 identical esp POE devices. the configs are all identical except the name and static IP address.
after update to 2025.11.0 I’m getting this error
“httpd httpd_accept_conn error in accept 23”

on devices that I did not update, I can click the refresh button as fast as i want and the page shows up every time.
after this new update, I get “this page isn’t working” and then a second or 2 later the page loads and the logs show the HTTPd error above.
I have tried to re-install the config and power cycled the device, but still see the same thing.

any ideas on a solution/cause?

2

Share your yaml.

3

I can post the yaml, but it’s the same one(s) I’ve been using for a year.
this issue didn’t occur until after i updated esphome to the latest version.
(running 11.5 now)
I turned on debug and this is what i see every time i click refresh.

> [01:07:54.546][D][web_server_idf:442]: Removing dead event source session
> [01:07:54.548][D][esp-idf:000][httpd]: E (66626) httpd: httpd_accept_conn: error in accept (23)
> [01:07:54.549][D][web_server_idf:533][httpd]: Event source connection closed (fd: 63)
> [01:07:58.066][D][esp-idf:000][httpd]: E (70158) httpd: httpd_accept_conn: error in accept (23)
> [01:07:58.105][D][web_server_idf:442]: Removing dead event source session
> [01:07:58.107][D][web_server_idf:533][httpd]: Event source connection closed (fd: 63)
> [01:08:01.085][D][web_server_idf:442]: Removing dead event source session
> [01:08:01.087][D][esp-idf:000][httpd]: E (73160) httpd: httpd_accept_conn: error in accept (23)
> [01:08:01.088][D][web_server_idf:533][httpd]: Event source connection closed (fd: 63)
4

OK - I can reproduce as well. Maybe you should log an issue on Github:

It doesn’t happen every refresh for me but it does happen:

[12:29:09.007][D][web_server_idf:442]: Removing dead event source session
[12:29:09.015][D][web_server_idf:533][httpd]: Event source connection closed (fd: 56)

Also:

[12:33:20.032][W][web_server_idf:588]: Closing stuck EventSource connection after 2500 failed sends
5

Thanks for validating. I will open an issue. I’m not updating my other hubs for now. it seems to be limited to the web server, as the rest of the device appears to work correctly.

6

I’m curious. Since the current version is 2025.11.5, is there a specific reason you chose to update to 2025.11.0?

I’m at 2025.11.2, and was trying to decide whether to update, or just wait for 2025.12.x. Obviously if there’s a problem with any of the 11.x versions, this would factor in to my decision. Thank you.

7

I’m on 2025.11.5 and experience the issue intermittently. The most noticeable symptom is that the web server log pane stops updating.

Doesn’t seem to affect anything else but web server.

8

I just updated to the latest, which was 11.5, home assistant said 11.0 was min. i opened a github issue, the reviewed and put in a fix, should be in the next release.

9

i opened a github issue, they identified the issue and put in a fix for the next release.

10

hi, have you the reference of git issue ? Same problem for me

11

A possible workaround that seems to work for me;

web_server:
local: true

12

I see also this issue on several ESP32 esphome devices.
Others increase CONFIG_LWIP_MAX_SOCKETS

see Multiple [httpd]: E (142012) httpd: httpd_accept_conn: error in accept (23) in project · Issue #12894 · esphome/esphome · GitHub

13

that’s probably the longest yaml config file i’ve ever seen… wow