ESP32 s2 mini P2P networking

[ilguvat]

HI
Finally managed to flash ESP32 s2 mini. The firmware ESP_Easy_mega_20241222_collection_A_ESP32s2_4M316k_CDC was installed. Tried several others, including factory - the board did not start.

I want to make S2 ​​mini a data collector from other different modules (ESP01, Wemos d1 mini, NodeMCU) via the ESPEasy P2P Networking protocol.
Currently, the Wemos d1 mini is acting as a collector. The problem is that it can only add 12 channels - in addition to its 4 physical sensors, there are only 8 from the P2P network. But we need more! In addition, the channels for adding sensors should not intersect for P2P devices, which further reduces the capabilities.

A question for the developers - Is it possible to implement adding sensors from P2P network devices using the Unit Name-Unit Number bunch, as is done, for example, when adding several D18B20 on OneWire.
So we could use, for example, channel 29 of the S2 mini to add channel 2 from the ESP01

[Ath]

A question for the developers - Is it possible to implement adding sensors from P2P network devices using the Unit Name-Unit Number bunch, as is done, for example, when adding several D18B20 on OneWire.
So we could use, for example, channel 29 of the S2 mini to add channel 2 from the ESP01

That is a planned enhancement, with the exact reasoning you have, but hasn't been implemented yet, mostly for time constraints, as it requires changes to internal data structures, and we aim for backward compatibility wherever possible, so older releases keep working as-is.

[TD-er]

A simple work-around (which isn't perfect I know) might be to send the values via rules.
So the sending task will trigger an event and then via rules you can then send a command to another node via p2p.

This also allows to make it more 'condensed' as you will be storing the values on the receiving end in a dummy task.
So for example you have 4 CO2 sensor tasks on some nodes which each deliver 1 (usable) value per task.
Using "taskValueSet" you can then store 4 of those in a single dummy task on the collecting node.

Code: Select all

on taskname#taskvaluename do
  SendTo,<unit nr>,"taskvalueset,<taskindex>,<taskvalueindex>,%eventvalue1%"
endon

You can make this quite compact when using things like "%eventpar%" and event wildcards.

Useful links:
https://espeasy.readthedocs.io/en/lates ... g-wildcard
https://espeasy.readthedocs.io/en/lates ... r-eventpar

[ilguvat]

Thanks for the quick feedback!
I'll try this method

Another small question not about p2p

When flashing the C2 mini, I noticed this oddity - After the first boot, when I enter my home WiFi network and the module receives an IP address (I see this in the router in the list of connected devices) - I cannot enter it at that address. But I noticed that if I save the NTP server in the advanced settings before saving the home network to the module - it enters by IP successfully.

What could this depend on!? Maybe in the address line after the IP I need to put / and add something?!

[TD-er]

I assume you're talking about the S2, not the C2...

What NTP-server do you set in the advanced settings?
An IP-address or a DNS hostname?
By default ESPEasy does not have "Use NTP" enabled on a clean install.
When only having the checkbox to use NTP checked, ESPEasy will try to reach one of the random hosts from pool.ntp.org.

So your ESPEasy node should be allowed to access the internet when you don't have a local NTP host configured in ESPEasy.

First it does a DNS lookup to pool.ntp.org and then tries to reach that host using UDP.

I know that DNS lookups can be blocking until you receive an answer.
Even though I do have set timeouts everywhere I can.

[ilguvat]

"When only having the checkbox to use NTP checked, ESPEasy will try to reach one of the random hosts from pool.ntp.org"

I've been using your firmware for 3-4 years now - NTP has never connected even to a random address!) The IP address and DNS are received from the router automatically.But I've never noticed such a problem, although maybe it's just me. I usually set up NTP manually right after connecting to my home network. I usually specify the least loaded server, like - ntp0.ntp-servers.net

Yes, I'm talking about c2, I changed the translator incorrectly.

Maybe it's worth adding a checkbox to the first network connection settings page with the NTP server settings?!

[TD-er]

Just to make sure there is no misunderstanding...
With "random server" I mean that pool.ntp.org will always return a randomly selected host from its pool.
You can try this yourself by trying to ping such a host via a command prompt on your PC.
You may need to wait for a few minutes between retries though as the DNS may be cached locally.

Every now and then such hosts may be down and when they are, it might cause a significant delay.
Also a NS lookup may sometimes stall for quite some time, especially when the connection is just active for a few (milli)seconds.
So I do have some delay added when the connection was made, just to prevent such extra waits.
The idea is that waiting a bit more is faster in the end.
However, this added delay is only based on what I've seen is usually a good compromise.

I will think about adding a NTP host to the IP-settings. The reason why I'm not immediately adding it there is that it doesn't feel that related and thus migh confuse users as they may feel the need to fill in the field, while it is not really needed.
On the other hand, it is technically possible to set a NTP server via DHCP, though I have not been able to access that info from a DHCP reply if it was set by the DHCP server.

[TD-er]

Ah nice example of pool.ntp.org not returning a working IP.

Code: Select all

C:\Users\gijsn>ping pool.ntp.org

Pinging pool.ntp.org [5.39.184.12] with 32 bytes of data:
Reply from 5.39.184.12: bytes=32 time=8ms TTL=58
Reply from 5.39.184.12: bytes=32 time=7ms TTL=58

Ping statistics for 5.39.184.12:
    Packets: Sent = 2, Received = 2, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
    Minimum = 7ms, Maximum = 8ms, Average = 7ms
Control-C
^C
C:\Users\gijsn>ping pool.ntp.org

Pinging pool.ntp.org [178.239.19.62] with 32 bytes of data:
Request timed out.
Request timed out.

Ping statistics for 178.239.19.62:
    Packets: Sent = 2, Received = 0, Lost = 2 (100% loss),
Control-C
^C

[ilguvat]

I have just conducted an experiment. The NTP server you specified in Windows gives 49ms. I entered it manually in the Advanced settings of S2 mini with the checkbox enabled - the time was also updated after rebooting the module with a 3-4 minute delay. But as soon as I made the checkbox inactive and rebooted with a 10 minute delay - the main page gave a message - Local Time: No system time source . And refused to synchronize with the default server.

Do I understand correctly - immediately after the ESP module contacts the Internet via the home network - the module itself should synchronize the time?! That is, in fact, the owner should only make a time zone correction. Correct!? Or did I misunderstand something!?

Here's another experiment - I disabled the checkbox and reset the power supply. Here's the boot log file

Code: Select all

644: Using PSRAM
647: ^^^INIT : Booting version: ESP_Easy_mega_20241222_collection_A_ESP32s2_4M316k_CDC, (GitHub Actions) mega-20241222 (ESP32 SDK 41
658: INIT : Free RAM:149052
670: INIT : Cold Boot - Restart Reason: CPU0: Vbat power on reset
672: FS   : Mounting...
680: FS   : SPIFFS mount successful, used 141815 bytes of 290156
692: CRC  : Settings CRC...OK
707: ESPEasy console using ESPEasySerial
729: ESPEasy console fallback enabled
738: CRC  : SecuritySettings CRC...OK 
746: Current Time Zone: STD time start: 1970-10-25 03:00:00 offset: 300 min
748: INIT : Check for Priority tasks
760: INIT : SPI not enabled
761: Set Network mode: WiFi
863: WIFI : Set WiFi to STA
1058: WiFi : Start network scan all channels
2758: WiFi : Scan finished, found: 1
2761: WiFi : Best AP candidate: Hexyu-cmotretb 46:D5:F2:26:11:26 Ch:1 (-50dBm) WPA2/PSK (bgn)
2773: WiFi : Added known candidate, try to connect
2776: WIFI : Set WiFi to OFF
2988: WIFI : Set WiFi to STA
3104: WIFI : Connecting Hexyu-cmotretb 46:D5:F2:26:11:26 Ch:1 (-50dBm) WPA2/PSK (bgn) attempt #0
3834: WIFI : Arduino wifi status: WL_CONNECTED 3 ESPeasy internal wifi status: DISCONNECTED
3836: WiFi : Best AP candidate: Hexyu-cmotretb 46:D5:F2:26:11:26 Ch:1 (-50dBm) WPA2/PSK (bgn)
3852: ESPEasy console using ESPEasySerial
3874: ESPEasy console fallback enabled
3875: INIT : Free RAM:112680
3891: ESPEasy console using ESPEasySerial
3913: ESPEasy console fallback enabled
3915: INFO : Plugins: 84 ['Normal','Collection_A ESP32'] (ESP32 SDK 41a885b-dirty)
3927: EVENT: System#Wake
3965: EVENT: System#BootMode=0,0,0,1,0,0
3969: Webserver: start
3973: EVENT: System#Boot
3981: WIFI : Connected! AP: Hexyu-cmotretb (46:D5:F2:26:11:26) Ch: 1 Duration: 654 ms
3995: WIFI : DHCP IP: 192.168.1.128 (Collection-A-ESP32) GW: 192.168.1.1 SN: 255.255.255.0 DNS: 192.168.1.1 / 0.0.0.0   duration: 20 
4007: Skip saving SecuritySettings, not changed
4025: UDP : Start listening on port 65500
4027: firstLoopConnectionsEstablished
4033: Internal command: All checked OK
4042: EVENT: WiFi#ChangedAccesspoint
4045: Webserver: stop
4048: Webserver: start
4053: EVENT: WiFi#ChangedWiFichannel
4058: EVENT: WiFi#Connected
5319: WD   : Uptime 0  ConnectFailures 0 FreeMem 102740 WiFiStatus: WL_CONNECTED 3 ESPeasy internal wifi status: Conn. IP Init
9337: static_file: /esp.css to: 192.168.1.238
35321: WD   : Uptime 1  ConnectFailures 0 FreeMem 104664 WiFiStatus: WL_CONNECTED 3 ESPeasy internal wifi status: Conn. IP Init
65308: WD   : Uptime 1  ConnectFailures 0 FreeMem 104764 WiFiStatus: WL_CONNECTED 3 ESPeasy internal wifi status: Conn. IP Init
95320: WD   : Uptime 2  ConnectFailures 0 FreeMem 104236 WiFiStatus: WL_CONNECTED 3 ESPeasy internal wifi status: Conn. IP Init
125319: WD   : Uptime 2  ConnectFailures 0 FreeMem 104236 WiFiStatus: WL_CONNECTED 3 ESPeasy internal wifi status: Conn. IP Init
155320: WD   : Uptime 3  ConnectFailures 0 FreeMem 104236 WiFiStatus: WL_CONNECTED 3 ESPeasy internal wifi status: Conn. IP Init

Next I turn on the checkbox and here are the changes

Code: Select all

155320: WD : Uptime 3 ConnectFailures 0 FreeMem 104236 WiFiStatus: WL_CONNECTED 3 ESPeasy internal wifi status: Conn. IP Init
185319: WD : Uptime 3 ConnectFailures 0 FreeMem 103708 WiFiStatus: WL_CONNECTED 3 ESPeasy internal wifi status: Conn. IP Init
215319: WD : Uptime 4 ConnectFailures 0 FreeMem 103708 WiFiStatus: WL_CONNECTED 3 ESPeasy internal wifi status: Conn. IP Init
245309: WD : Uptime 4 ConnectFailures 0 FreeMem 103708 WiFiStatus: WL_CONNECTED 3 ESPeasy internal wifi status: Conn. IP Init
275474: WD : Uptime 5 ConnectFailures 0 FreeMem 103132 WiFiStatus: WL_CONNECTED 3 ESPeasy internal wifi status: Conn. IP Init
280599: ESPEasy console using ESPEasySerial
280621: ESPEasy console fallback enabled
280627: SaveToFile: free stack: 6820
280675: FILE : Saved config.dat offset: 0 size: 3028
280676: SaveToFile: free stack after: 6820
280688: Skip saving SecuritySettings, not changed
281135: NTP : NTP replied: delay 195 ms round-trip delay: 196 ms offset: 20200T14:29:18.134,^ t0: 1745331397.579,^ t1: 280.925,^ t2
281138: Time set to 1745332439.257
281150: Current Time Zone: STD time start: 2025-10-26 03:00:00 offset: 300 min
281152: Local time: 2025-04-22 19:33:59
281495: EVENT: Time#Initialized
281583: EVENT: Clock#Time=Tue,19:33
282754: EVENT: Clock#Time=Tue,19:34
305320: WD : Uptime 5 ConnectFailures 0 FreeMem 102492 WiFiStatus: WL_CONNECTED 3 ESPeasy internal wifi status: Conn. IP Init

And another observation - when the checkbox is disabled, the processor load is about 68%

Code: Select all

Local Time:	No system time source
Time Source:	No time set
Uptime:	00h00m
Load:	68.04 [%] (LC=363)
Internal Temperature:	41.30 [°C]

And as soon as I activate the checkbox and the time is synchronized, the load becomes about 11%

Code: Select all

Local Time:	2025-04-22 19:38:48
Time Source:	NTP
Uptime:	00h02m
Load:	11.75 [%] (LC=5466)
Internal Temperature:	41.47 [°C]

Here is another screenshot of the ping. It's strange, 5-6 hours ago it was working fine. Now it gives an error
https://ibb.co/99YXr7zd

[TD-er]

The CPU-load should not be affected by this.
Can you on the tools->Advanced page enable the "Collect Timing Statistics" checkbox?

Then after a while you can visit the Tools-> Timing Stats page.

N.B. on each reload of that page, the stats will be reset.

This may help you find why your node was running so slow.
A loop-count (LC) of 363 is not normal, unless you have ECO mode enabled.

And maybe also update to the latest build, which will probably be the next official build: https://td-er.nl/ESPEasy/latest/

[ilguvat]

Can you on the tools->Advanced page enable the "Collect Timing Statistics" checkbox?
Then after a while you can visit the Tools-> Timing Stats page.
I turned it on, but I don't understand much here. I'll attach a screenshot below.
https://i.postimg.cc/pVwmmRF3/del.jpg

A loop-count (LC) of 363 is not normal, unless you have ECO mode enabled.
Perhaps this is because this module is empty for now - nothing is physically connected to it.
Now I remotely looked at the LC value on the working WeMos D1 mini module - about 9000

https://td-er.nl/ESPEasy/latest/
The link does not open even through the Dutch VNP. Maybe there is a mirror of the site, or on GitHub

[Ath]

I turned it on, but I don't understand much here. I'll attach a screenshot below.
https://i.postimg.cc/pVwmmRF3/del.jpg

Can you attach images like that in the Forum Attachment tab? (Below the message edit area) External image hosters tend to remove these images after a couple of weeks/months, making your information 'less useful'.

For quoting previous messages, there's a Quote button available (the big double-quote in the right-top of each message body)

[ilguvat]

Can you attach images

Thanks for your help.
Edited the message above

[TD-er]

[...]
https://td-er.nl/ESPEasy/latest/
The link does not open even through the Dutch VNP. Maybe there is a mirror of the site, or on GitHub

Can you try again?
I do have some rather extensive filter on it to try to block hack attempts as I do have a few TByte/month of incoming data which is being blocked.

If it still doesn't work, can you send me the time you tried so I can try to find your IP from the logs and remove your IP-block (if it is a block on my server).
Or you can send me a private message with your IP.

Anyway, the same build is here: https://github.com/letscontrolit/ESPEas ... 4604630407

[ilguvat]

build is here: https://github.com/letscontrolit/ESPEas ... 4604630407

Updated firmware from web interface

del (1).jpg (129.39 KiB) Viewed 1092 times

uild: ESP_Easy_mega_20250422_climate_ESP32s2_4M316k_CDC Apr 22 2025

del.jpg (262.68 KiB) Viewed 1092 times

If you need to try to flash a specific one, I can do it. I have time for this.

[TD-er]

Hmm given the time it takes for getting a DNS name resolved and your NTP duration, I think you have a rather slow internet connection. (or very busy connection)

So keep this in mind when you need to configure something that's not in your local network, to increase any used timeout.
For example when configuring a MQTT broker which is online and not in your local network.

There seems to be only a single call to the NTP server, so it isn't like ESPEasy is trying continuously to get the time from the NTP server.

[ilguvat]

That's right. The connection where I am now is not very good at the moment.

About MQTT, here's another question, since we've already touched on this topic...
Most of my sensors work via the MQTT protocol + transmit readings to the NarodMon.ru service (configured via Rules). Periodically, the modem freezes and I reboot it by disconnecting the power supply - after that, the connection appears, the data is transmitted to the service NarodMon. But MQTT does not connect to the north until I reboot it programmatically
Two modules with firmware from 2021 and 2022 behave this way. I'm afraid to update - I'm afraid I won't be able to configure it as it was))

[TD-er]

You can save the current settings file (and rules) and keep them as a backup.
This is always a good idea to do.
On later builds you even can download all files as a single tar file, but that wasn't an option back in 2022, so you may need to save all files on your PC.

Please save those files before making any update, as newer versions may apply transitions to the settings file as the internal structures may have changed.

When updating to a newer version and you do have the files saved, I suggest you go for a "LittleFS" build as this is using a much newer SDK.
If your board shows poor WiFi/network performance after the update, you may need to call the command

Code: Select all

EraseSDKwifi

via the serial console or enter it via the command option on the tools page.
This will then erase any existing WiFi RF calibration and perform the calibration again.
It may take a few minutes before the ESP will feel 'snappy' again.
Make sure you do this while having access to the ESP as it will loose connection when doing this and you may need to power cycle it.

For links to the latest builds I just uploaded, see: https://github.com/letscontrolit/ESPEas ... 2824598958

[ilguvat]

I tried to flash NodeMCU 8266 and VeMos D1 mini for new firmware that was not officially released (Build: ESP_Easy_mega_20250422_normal_ESP8266_4M1M) - I was able to successfully flash only version, after flashing the rest, both modules did not create an access point for configuration. And in this version, both modules get an error when saving any changes (screenshot)

del.jpg (99.58 KiB) Viewed 936 times

Then I remotely saved the configuration of one of the working Wemos D1 mini modules with the old firmware I mentioned above (version 2022) and tried to load it into the same Wemos module, which I had just flashed with the new firmware - as a result, the new module wrote that the configuration was successfully saved, it was necessary to reboot. I reboot - and it again creates an access point and asks to connect to the home network. However, it did not transfer any configuration settings. It also resets Wi-Fi settings and creates its own access point during a software reboot from the VEB interface

del (2).jpg (130.39 KiB) Viewed 936 times

del (1).jpg (96.39 KiB) Viewed 936 times

When trying to update the firmware from the WEB interface, it gives an error - Update error: ERROR[2]: Flash Erase Failed
I wanted to try the Climate firmware

[Ath]

You may have done too many saves on a single day. Power-cycling the unit should solve it for now. Or there isn't enough free space to save the file. If that doesn't solve it, then you may have found the end of life for the flash chip

[ilguvat]

Yes, I know that this can be solved by rebooting the module, but here, immediately after rebooting, when saving the first change, such an error appeared.

Version 1m was installed via cable normally, but not via ESP Easy software - as it gave the same error 2
I used the program WIFI-IOT Flasher from the site wifi-iot.com.

I installed version 4M1M because I had already installed this version on these modules, but in 21-22-23 release years. Also, ESP Easy Flasher gave error 2 when selecting version 1M.
And then for some reason the module began to resist and glitch with version 4M1M.

[TD-er]

Can you show the flash section of the sysinfo page?
Maybe also run the command "meminfodetail" first on the tools page and then load the sysinfo page.

This command will enable drawing some charts at the end of the sysinfo page regarding the partition table.

[TD-er]

[...]
Version 1m was installed via cable normally, but not via ESP Easy software - as it gave the same error 2
I used the program WIFI-IOT Flasher from the site wifi-iot.com.

I installed version 4M1M because I had already installed this version on these modules, but in 21-22-23 release years. Also, ESP Easy Flasher gave error 2 when selecting version 1M.
And then for some reason the module began to resist and glitch with version 4M1M.

I have never seen this site, so I have no idea what this does.
I do know that some other firmwares, other than ESPeasy, may set a different partition layout.
So maybe your partition layout is set to force a too small file system?
Updating ESPEasy will not change the partition layout.

[TD-er]

Just another thought about the reported NTP issues...
Do you have IPv6 enabled (or even included in your build) ?
If it is enabled, you could see the actual active DNS servers on the sysinfo page.
Could you have a look at the sysinfo page to see what ESPEasy lists there as DNS servers?