Receive "scan_evt timeout" on serial monitor

[sami1369]

Hi guys
I use an ESP32 as scanner and 2 ESP32 Advertised to Scanner
َAfter A few hours I receive a message on serial monitor in scanner device and scanner doesn't work while other peripherals work perfect.
message is "scan_evt timeout"
my code in scanner:

Sketch:

void Task1code(void *pvParameters)
{
  for (;;)
  {
    BLEScanResults foundDevices = pBLEScan->start(scanTime, true);
    allowScan = true;
    Serial.println("Scan done!");
    pBLEScan->clearResults(); // delete results fromBLEScan buffer to release memory

    vTaskDelay(5);
  }
}

scan time is 3 second.

can you help me?

[juanaviladev]

Hi! +1, same here, has anyone found any solution to this? I've tried v2.0.2 and can still reproduce it.

Thanks in advance!

[beats0126]

same meet this issues after updating to esp32-arduino v2.0.2

[juanaviladev]

same meet this issues after updating to esp32-arduino v2.0.2

Tried using ESP-IDF v4.3.1 sdk (without the arduino layer) and still able to reproduce it, looks like an ESP-IDF problem with VHCI. espressif/esp-idf#4001

Anyway, has anyone found any way to handle the error? It's logged in the serial output, but I haven't found any callbacks or events that can be used to restart if "scan_evt timeout" occurs.

[stevenli2020]

i had the same "scan_evt timeout" issue, been scratching my head for the past 2 weeks. did anyone managed to find a fix?

[stevenli2020]

currently i am testing adding a delay of 10ms before "pBLEScan->clearResults();" - suspect clearResults happens too fast while callback function is still retrieving data. not sure if this could help...
the test has been going on for 2 days now, so far still running ok

[stevenli2020]

i have bad news and good news.
bad news - the test failed after continuous scanning for about 60 hours, with the same Scan_evt Timeout error and everything fell into a deadlock. so adding the delay did not solve the problem.
however, good news is - i have successfully force the board to reboot when this error happens, by using Task Watchdog Timer (TWDT) .
here are the code fragments -

#include <esp_task_wdt.h>
..
int _T;
int scanTime = 2; //In seconds
int scanInterval = 10; //In mili-seconds
void setup() {
...
  esp_task_wdt_init(3, true); //enable panic so ESP32 restarts, interrupt when task executed for more than 3 secons
  esp_task_wdt_add(NULL); //add current thread to WDT watch 
  _T = millis(); 
...
}
void loop() {
  for(int i=0; i<1000; i++){
    Serial.printf("<%d>\n",i); 
    delay(scanInterval);
    if(!scanBLE()){
      initBLEScan();
      esp_task_wdt_reset(); // reset the watchdog timer
    }
    checkSerialCmd();
    checkWifi();
    digitalWrite(LED_BUILTIN, LOW);   
  }
  initBLEScan();
}

In the above codes, the scanning time is set to 2 seconds and scan interval 10ms, the tested time between each scan is around 2040ms, therefore the TWDT is set to 3 seconds - when scan_evt timeout occures, if hangs for more than 3 seconds, the TWDT will reboot the device.
After running this codes over night, it worked!
below are the logs i have captured when the scan_evt timeout happens -

...

<626> Publishing > 0d:b297:0d27:0072
Interval=2064, Scan done! Devices found: 19
<627> Interval=2038, Scan done! Devices found: 24
<628> Interval=2041, Scan done! Devices found: 19
<629> Publishing > 0d:b298:0d23:0074
Interval=2040, Scan done! Devices found: 17
<630> Interval=2040, Scan done! Devices found: 21
<631> Interval=2039, Scan done! Devices found: 21
<632> Publishing > 0d:b299:0d2a:0071
Interval=2071, Scan done! Devices found: 18
scan_evt timeout
<633> E (19766618) task_wdt: Task watchdog got triggered. The following tasks did not reset the watchdog in time:
E (19766618) task_wdt:  - loopTask (CPU 1)
E (19766618) task_wdt: Tasks currently running:
E (19766618) task_wdt: CPU 0: IDLE
E (19766618) task_wdt: CPU 1: IDLE
E (19766618) task_wdt: Aborting.

abort() was called at PC 0x400e33cd on core 0


Backtrace:0x400834dd:0x3ffbe9ac |<-CORRUPTED




ELF file SHA256: 0000000000000000

Rebooting...
ets Jun  8 2016 00:22:57

rst:0xc (SW_CPU_RESET),boot:0x1b (SPI_FAST_FLASH_BOOT)
configsip: 0, SPIWP:0xee
clk_drv:0x00,q_drv:0x00,d_drv:0x00,cs0_drv:0x00,hd_drv:0x00,wp_drv:0x00
mode:DIO, clock div:1
load:0x3fff0030,len:1184
load:0x40078000,len:12804
ho 0 tail 12 room 4
load:0x40080400,len:3032
entry 0x400805e4

New bootup [1097BD1A8454]

MQTT - waiting for connection.


Connecting to WiFi .....
MQTT - waiting for connection.
192.168.3.14
Scan initializing ...
<0> Interval=2018, Scan done! Devices found: 17
<1> Publishing > 0d:b29b:0d21:0072
Interval=2065, Scan done! Devices found: 22
<2> Interval=2049, Scan done! Devices found: 20

...

the entire recovery took about 5s before the ESP32 is up and running again, this walk around is not perfect solution but works for my project.
hope it helps!

[VojtechBartoska]

Hello@stevenli2020, thanks for giving it a shot. Did you run it on 2.0.3-RC1?

[stevenli2020]

Hello@stevenli2020, thanks for giving it a shot. Did you run it on 2.0.3-RC1?

I have not test it on 2.0.3-RC1, my boards are FireBeetle ESP32. however i believe the TWDT will still be supported on the 2.0.3-RC1, isnt it?

[Whoswatching89]

Still running into this issue tried your fix @stevenli2020 getting error 'scanBLE' was not declared in this scope.

[thorrak]

I’ve been receiving this issue as well - though via h2zero’s amazing NimBLE port for Arduino. The Bluetooth stack stops responding every few hours (though in some cases within 15 minutes, and in others lasting 8 hours+). This is with the latest Arduino framework, built upon esp-idf v4.4.4.