Fixed 6h reboot on networks that filter the ESP32 outbound traffic

http_request runs on the loop task, which has a 4s budget before the
5s task WDT panics. A dropped connect() blocks for 4.5s (the ESPHome
default timeout) and a filtered port 53 blocks for a fixed 7s, which
timeout does not cover.

- timeout 3s and watchdog_timeout 60s cover the two blocking paths
- a failure counter stops the service after 3 network-level failures,
  instead of freezing the loop 14s every 6h forever
- the POST is now skipped when the GET failed

Reproduced and measured on OPNsense.
See Design_decisions.md.
This commit is contained in:
Sleeper85 committed 2026-11-08 23:08:33 +01:00
1 parent 9f90436849
commit e4b8aacc0a
4 files changed
+99 -21

No files matched your search

+46
View File
@@ -113,6 +113,52 @@ interval to make protocols "consistent" with each other.
When adding a frame to a sequence, remember that the sequence length is tracked alongside the
array. Keep the two in sync, or the new frame will silently never be emitted.
## The update service must never block the loop task
**File:** `packages/yambms/yambms_service.yaml`
`http_request` runs synchronously on the ESP32 loop task. ESPHome arms the IDF task watchdog at
5s with `CONFIG_ESP_TASK_WDT_PANIC=y`, and feeds it at most every 1 s, so any single blocking
call has a **4s budget** before the device panics and reboots.
Two paths exceed that budget on a network that filters outbound traffic. Both were reproduced on
an OPNsense firewall and measured on hardware:
| Path | Blocking time | Bounded by |
|---|---|---|
| Port 80 dropped silently | 4.5s | `timeout`, whose ESPHome default is 4.5s |
| Port 53 filtered (drop **or** reject) | 7.0 s, fixed | nothing — `timeout` does not cover DNS |
The 7s figure is the lwIP resolver retry ladder (`1 + 1 + 2 + 3`s over `DNS_MAX_RETRIES`), and
it is deterministic to a few milliseconds. A reject is not faster than a drop: lwIP's UDP API has
no error callback, so the ICMP unreachable is counted and discarded, and the query still runs its
full budget. Only an actual DNS *answer* — including `NXDOMAIN` from a local resolver — fails fast.
This is why three separate settings are needed, and none of them is redundant:
- **`watchdog_timeout: 60s`** is the only thing that covers the DNS wait. It is never reached in
normal operation, where a full request takes under 400 ms.
- **`timeout: 3s`** brings the socket phase back under the 4s budget. It does **not** help the
DNS case.
- **The failure counter** (`yambms${yambms_id}_var_service_failures`) is what stops the loop from
freezing for 7s every 6h, forever, on a device that will never reach the internet.
`my_network.is_connected()` is not a reachability test — it only proves the device holds an IP.
It is kept as a cheap first filter, but the counter is the real guard. Do not remove it on the
grounds that the connectivity check above it "already handles this".
The counter is incremented from `on_error`, which fires only when no connection could be made.
An HTTP error status still fires `on_response` and resets the counter, which is correct: the
network worked, so there is no freeze to protect against.
The POST is guarded by an `if` on the counter rather than nested inside the GET's `on_response`.
Nesting would stack one blocking request inside another on a loop task whose stack is 8192 B by
default — see [LOOP_TASK_STACK.md](../LOOP_TASK_STACK.md). It also avoids a second 7s freeze on
a filtered network, halving the worst case.
Reaching 3 failures disables the service until reboot, or until the user toggles the
`Update service` switch off and on, which rearms the counter.
## Empirical constants
Some constants were found by trial on real hardware and must not be rounded or harmonised.
+4
View File
@@ -240,6 +240,10 @@ globals:
type: bool
restore_value: no
initial_value: 'true'
- id: yambms${yambms_id}_var_service_failures
type: int
restore_value: no
initial_value: '0'
- id: yambms${yambms_id}_var_initialized
type: bool
restore_value: no
+43 -19
View File
@@ -1,5 +1,5 @@
# Updated : 2026.08.14
# Version : 1.2.9
# Updated : 2026.08.21
# Version : 1.3.0
# GitHub : https://github.com/Sleeper85/esphome-yambms
# YamBMS ( Yet another multi-BMS Merging Solution )
@@ -46,6 +46,14 @@ text_sensor:
# https://esphome.io/components/http_request.html
http_request:
verify_ssl: false
# Must stay under the ESP32 task WDT budget: a 5s WDT minus the 1s feed
# interval leaves 4s. The ESPHome default of 4.5s exceeds it and panics the
# loop task whenever a connect() is silently dropped by a firewall.
timeout: 3s
# DNS resolution runs before the socket and is NOT covered by `timeout`. A
# filtered port 53 blocks the loop task for a fixed 7s, which panics the WDT.
# See Design_decisions.md.
watchdog_timeout: 60s
interval:
- interval: 6h # 60min
@@ -54,33 +62,49 @@ interval:
condition:
and:
- lambda : return id(yambms${yambms_id}_var_service) == true;
- lambda : return id(my_network).is_connected();
- lambda : return id(my_network).is_connected(); # only proves the device has an IP, not that the internet is reachable.
- lambda : return id(yambms${yambms_id}_var_service_failures) < 3;
then:
- http_request.get:
url: http://script.opentel.be/yambms.get.php
capture_response: true
# on_error fires only when no connection could be made (DNS or
# connect failure) — the cases that freeze the loop. An HTTP error
# status still fires on_response, which rightly resets the counter.
on_error:
then:
- lambda: id(yambms${yambms_id}_var_service_failures) += 1;
on_response:
then:
- lambda: |-
id(yambms${yambms_id}_var_service_failures) = 0;
json::parse_json(body, [](JsonObject root) -> bool {
// Explicitly cast to std::string to resolve ambiguity
id(yambms${yambms_id}_last_version).publish_state(root["version"].as<std::string>());
return true;
});
- http_request.post:
url: http://script.opentel.be/yambms.post.php
request_headers:
Content-Type: application/json
json: |-
root["mac"] = id(esp32_mac_address).state;
root["version"] = "${yambms_version}";
root["board_chip"] = "${board_chip}";
root["board_name"] = "${board_name}";
root["bms_model"] = id(yambms${yambms_id}_var_bms_models);
root["bms_protocol"] = id(yambms${yambms_id}_var_bms_protocols);
root["bms_count"] = id(yambms${yambms_id}_var_bms_counter);
root["comm_bus"] = id(yambms${yambms_id}_var_communication_bus);
root["comm_protocol"] = id(yambms${yambms_id}_var_communication_protocol);
root["uptime"] = id(esp32_uptime).state;
root["date_time_now"] = id(esp32_date_time_now).state;
# Guarded by an `if` rather than nested in on_response: nesting would
# stack a blocking request inside another one on the loop task, whose
# stack is 8192 B by default (see LOOP_TASK_STACK.md). On a filtered
# network it also avoids a second 7 s freeze for nothing.
- if:
condition:
lambda: return id(yambms${yambms_id}_var_service_failures) == 0;
then:
- http_request.post:
url: http://script.opentel.be/yambms.post.php
request_headers:
Content-Type: application/json
json: |-
root["mac"] = id(esp32_mac_address).state;
root["version"] = "${yambms_version}";
root["board_chip"] = "${board_chip}";
root["board_name"] = "${board_name}";
root["bms_model"] = id(yambms${yambms_id}_var_bms_models);
root["bms_protocol"] = id(yambms${yambms_id}_var_bms_protocols);
root["bms_count"] = id(yambms${yambms_id}_var_bms_counter);
root["comm_bus"] = id(yambms${yambms_id}_var_communication_bus);
root["comm_protocol"] = id(yambms${yambms_id}_var_communication_protocol);
root["uptime"] = id(esp32_uptime).state;
root["date_time_now"] = id(esp32_date_time_now).state;
+6 -2
View File
@@ -1,5 +1,5 @@
# Updated : 2026.06.27
# Version : 1.1.6
# Updated : 2026.08.21
# Version : 1.1.7
# GitHub : https://github.com/Sleeper85/esphome-yambms
# YamBMS ( Yet another multi-BMS Merging Solution )
@@ -49,6 +49,10 @@ switch:
- globals.set:
id: yambms${yambms_id}_var_service
value: 'true'
# Rearm the circuit breaker, so the user can retry after fixing the network
- globals.set:
id: yambms${yambms_id}_var_service_failures
value: '0'
turn_off_action:
- globals.set:
id: yambms${yambms_id}_var_service