Testers for new upcoming build mid August '23

[TD-er]

In the past, some users (like @fly74 ) have been reporting severe issues only hours after a new build.
So to prevent such mishaps and perhaps even act like we're professionals developing software, I would like to invite people to do some testing on the latest code base.

I also copied the latest GHActions test build files on the web flasher page: https://td-er.nl/ESPEasy/latest/ (2023-08-13)

ToDo:

• Crash when not checking all 'Stats' checkboxes for CSE7766 (at least 1 has to be checked) (merged)
• I2C address shift for P113 - VL53L1X (Merged)
• Cannot change default taskvaluenames for Eastron plugin. (Still needs a fix)

[fly74]

(like @fly74 )

Still in vacation.

[TD-er]

(like @fly74 )

Still in vacation.

So at least you won't report any serious issues without hours of the new build :)

[iz8mbw]

Using latest mega_a48dc65 build on 2 ESP32 boards, one Climate and one Energy builds.
No issues for me

[ghtester]

I am running various custom firmware releases. The bad thing on latest is a huge memory leak as discussed in #4726. This looks to be an issue with the MDNS feature. The custom firmware dated 20230711, compiled with MDNS disabled + PR4354 is significantly more stable - even though I have encountered a crash of one node running this version, it looks useful for me. So I vote for temporary disabling the MDNS until a memory leak is fixed.
As I have a customized firmware, I can't use the test build but I could make a fresh custom build with recent updates. Nevertheless, the stability testing takes a very long time so it's perhaps irrelevant.

BTW. Thanks to all developers & contributors for their great work!

[HV-NL]

Installed mega_a48dc65 on 2 ESP8266 boards: display 4M (1.8" ST77xx TFT module) and Climate (DHT22 sensor).
No issuse seen.

[sp2ong]

Hi I try latest build 20230808 with ESP8266 Normal build, but I don't see futures which I tested

#4722

Optional user-defined display headers will be added in release 20230823 ?

[fly74]

(like @fly74 )

Still in vacation.

So at least you won't report any serious issues without hours of the new build :)

@TD-er No. If I want I could flash remotely. But it is not a good idea for testing. 👀

[iz8mbw]

@fly74 I think you can flash remotely without any issues

[TD-er]

Well I do think it is still a good idea to not perform an OTA update from a holiday location :)

[TD-er]

Hi I try latest build 20230808 with ESP8266 Normal build, but I don't see futures which I tested

#4722

Optional user-defined display headers will be added in release 20230823 ?

That one hasn't been merged yet.
I will discuss with Ton whether it is a good idea to merge it before the next build.

[ghtester]

Just one thing to point out - as already mentioned in another thread ( #4726 (comment) ), I have experienced an issue with MQTT controller which couldn't reconnect to MQTT broker (configured with host name) after temporary connectivity lost (but WiFi connectivity was not interrupted). This happened again, custom build dated 20230711.

[TD-er]

That's already with sources of 4 weeks ago....

[fly74]

@TD-er Back on duty. I tested

├── ESP32-D0WDQ6_neopixel_ESP32_4M316k
│   └── ESP_Easy_mega_20230808_neopixel_ESP32_4M316k.bin
├── ESP32-D0WDQ6_normal_ESP32_4M316k
│   └── ESP_Easy_mega_20230808_normal_ESP32_4M316k.bin
├── ESP8266_climate_ESP8266_4M1M
│   └── ESP_Easy_mega_20230808_climate_ESP8266_4M1M.bin
├── ESP8266_neopixel_ESP8266_4M1M
│   └── ESP_Easy_mega_20230808_neopixel_ESP8266_4M1M.bin
└── ESP8266_normal_ESP8266_4M1M
    └── ESP_Easy_mega_20230808_normal_ESP8266_4M1M.bin

No issues at all.

[TD-er]

Thanks for reporting for duty :)

There have been 3 issues found, which will be fixed for the next test build:

• Crash when not checking all 'Stats' checkboxes for CSE7766 (at least 1 has to be checked) (fixed, not yet merged)
• I2C address shift for P113 - VL53L1X (PR is ready to be tested)
• Cannot change default taskvaluenames for Eastron plugin. (Still needs a fix)

[TD-er]

Updated to the latest build (see start post)

[fly74]

No problems so far with:

├───ESP32-D0WDQ6_neopixel_ESP32_4M316k
│       ESP_Easy_mega_20230813_neopixel_ESP32_4M316k.bin
│
├───ESP32-D0WDQ6_normal_ESP32_4M316k
│       ESP_Easy_mega_20230813_normal_ESP32_4M316k.bin
│
├───ESP8266_climate_ESP8266_4M1M
│       ESP_Easy_mega_20230813_climate_ESP8266_4M1M.bin
│
├───ESP8266_neopixel_ESP8266_4M1M
│       ESP_Easy_mega_20230813_neopixel_ESP8266_4M1M.bin
│
└───ESP8266_normal_ESP8266_4M1M
        ESP_Easy_mega_20230813_normal_ESP8266_4M1M.bin

[iz8mbw]

All OK here using mega_c25341d build with:

ESP32 - Energy
ESP32 - Climate
ESP32 - Max

[ghtester]

Just to mention a strange high Exceptions rate on a node with (almost) empty configuration - upgraded from a couple months old firmware to custom build dated 20230808. There are no plugins configured, no controllers configured, no rules, it's only connected to a stable WiFi AP.

Unit Number:	0
Local Time:	2023-08-17 20:11:24
Time Source:	NTP
Time Wander:	41.0 [ppm]
Uptime:	3 days 2 hours 55 minutes
Load:	3.24% (LC=12443)
CPU Eco Mode:	false
Boot:	Exception (41)
Reset Reason:	Exception
Last Action before Reboot:	Background Task
SW WD count:	0

Firmware

Build:⋄ | 20230808 - Mega
System Libraries:⋄ | ESP82xx Core 2843a5ac, NONOS SDK 2.2.2-dev(38a443e), LWIP: 2.1.2 PUYA support
Git Build:⋄ | My Build: Aug 8 2023 14:59:15
Plugin Count:⋄ | 40 [IR][No Debug Log]
Build Origin: | Vagrant
Build Time:⋄ | Aug 8 2023 14:58:55
Binary Filename:⋄ | ESP_Easy_mega_20230808_custom_IR_ESP8266_4M1M
Build Platform:⋄ | Linux-4.15.0-51-generic-x86_64-with-glibc2.27
Git HEAD:⋄ | mega_a48dc65

[TD-er]

It is already running without a reboot for 3 days, so I find it strange to call it a "high exception rate"
Sure any exception reboot is one too many, unless you've got MDNS enabled as that seems to have a memory leak.
And memory leaks will result in a crash due to memory allocation -> Exception.

[ghtester]

Yes it was working without a reboot for last 3 days but that really means a "high exceptions rate" during the days before. ;-)
I was not observing / monitoring that node as I did not expect so high crashes count and it was a surprise to me... I wonder the reason, I have no idea why this happened while the node was in fact in idle state... To be observed...

[iz8mbw]

I have 3 ESP32 nodes with an uptime more than 4 days without any "Boot Exception".

[TD-er]

I just uploaded this latest GitHub Actions build
to the latest web flasher: https://td-er.nl/ESPEasy/latest/

I intend to make a new build based on this last test build.
So please let me know if there's something with this build.

The only intended change we'll make before the next build is to update the Espressif download tool and remove the ESPEasy flasher application as this only supports a mere fraction of all the boards we now support (the old flasher only supports ESP8266 and only for a limited subset of USB to serial chips)
And the other features offered by this are now also supported via the web flasher, now that we support the IMPROV protocol.

[iz8mbw]

@TD-er
Just installed latest mega_8c3f67f build on 3 ESP32 boards: one Climate, one Energy and one Max.
No issues for me

[TD-er]

A new build is made and uploaded to the releases page and the web flasher (one may need to press CTRL-F5 for refresh of the index.html page)
So this can be closed.

Thanks you all for testing.