Cumulus MX V4 Build Webtags Not working.

[nicksnet]

Hi Mark,

Yes I did change that value, same result, not working, also added Airlink webtags to CUtags.php, after re-writing tags to the same configuaration as the original tags, again not working, its as if the Airlink tags are not even there, even though they are when checking Cumulus settings etc.

Anyway, I'm putting this one to bed for the time being, have changed banner to show Solar Radiation, which indeed it does, as my station is testing at present the reading is zero as expected, which is the tag value.

Thanks again Mark, but enough time spent on this one, maybe a future look might be worth it, once airlink is setup outside.

Nick.

[BeaumarisWX]

Hi Mark,

Just upgraded a few hours ago to Build 4039 and seeing the following returns :

Code: Select all

"#NewBuildAvailable"         = "n/a";  //  value of 1 if a New Build is available, 0 if not

"#NewBuildNumber"           = "n/a";  //  returns the latest available Cumulus MX build number

https://beaumaris-weather.com/CUtags.php?sce=dump

Kind Regards,

[mcrossley]

They are working for me.

Code: Select all

http://<IP_ADDRESS>:8998/api/tags/process.json?NewBuildAvailable&NewBuildNumber

{
    "NewBuildAvailable": "0",
    "NewBuildNumber": "4039"
}

Was the network available when Cumulus started?

The "n/a" would indicate it could not contact GitHub to retrieve the latest version information at start-up.

The version info will refresh once a day. So, I expect the tags will work for you again sometime tomorrow.

[BeaumarisWX]

Hi Mark,

As usual you are correct

Investigation proved a network issue at new version startup time, all back to normal, thanks.

Code: Select all

{
	"NewBuildAvailable":"0",
	"NewBuildNumber":"0000"
}

kind Regards,

[BeaumarisWX]

OK, so this still does not work for me, sorry Mark.

I purposely have not upgraded to new version 4.2.1 (build 4040) still on 4.2.0 (build 4039).

The api call and the processed tags (uploaded to my website etc.) still return ("{"NewBuildAvailable":"0","NewBuildNumber":"0000"}").

It should return ("{"NewBuildAvailable":"1","NewBuildNumber":"4040"}").

I have also purposely rebooted CMX and it is saying in the cmd window and the LatestError tag states "You are not running the latest version of Cumulus MX, build Build 4040 - Version 4.2.1 is available" < (this part is correct).

Additional info : I awoke this morning to see that the "#UpgradeAlarm" was Active (1, not 0) in the dashboard footer and that the Tag "#UpgradeAlarm" processed and uploaded to my website was also Active (1, not 0) though as mentioned above ("{"NewBuildAvailable":"0","NewBuildNumber":"0000"}").

That was when I decided (many hours later) to not update new version and just reboot CMX, since then "#UpgradeAlarm" has been in-Active (0, not 1) in the dashboard footer and that the Tag "#UpgradeAlarm" processed and uploaded to my website was also in-Active (0, not 1) been hours now and no change ?

Can anyone else confirm this issue ?

Kindest Regards,

[freddie]

Can anyone else confirm this issue ?

Yes, I see this behaviour too. When I first checked (07:10 BST) I hadn't had the alarm notification of a new release. This was triggered at 07:26 BST. The tag output remained the same after the alarm was triggered.

I noticed some other behaviour too, which I think is new. When the alarm was triggered it also triggered a restart of MX. It could be that something else triggered a restart of MX, and the alarm was triggered as part of the restart. I don't know.

Diags files attached.

Excerpt from syslog:

Code: Select all

2024-10-16T06:26:19.975454+00:00 weather dotnet[90891]: Cumulus stopped
2024-10-16T06:26:20.008809+00:00 weather systemd[1]: cumulusmx.service: Main process exited, code=exited, status=143/n/a
2024-10-16T06:26:20.009285+00:00 weather systemd[1]: cumulusmx.service: Failed with result 'exit-code'.
2024-10-16T06:26:20.009597+00:00 weather systemd[1]: Stopped cumulusmx.service - CumulusMX service.
2024-10-16T06:26:20.009781+00:00 weather systemd[1]: cumulusmx.service: Consumed 11h 47min 29.797s CPU time, 180.2M memory peak, 0B memory swap peak.
2024-10-16T06:26:20.021521+00:00 weather systemd[1]: Started cumulusmx.service - CumulusMX service.
2024-10-16T06:26:21.202345+00:00 weather dotnet[115674]: Cumulus running at: http://localhost:8998/
2024-10-16T06:26:23.005757+00:00 weather dotnet[115674]: 10/16/2024 06:26:23
2024-10-16T06:26:23.005887+00:00 weather dotnet[115674]: Type Ctrl-C to terminate
2024-10-16T06:26:23.011094+00:00 weather dotnet[115674]: Downloading Archive Data
2024-10-16T06:26:23.222451+00:00 weather dotnet[115674]: Downloading Archive Data

20241016-053000.txt

20241016-062620.txt

[BeaumarisWX]

Hi Freddie,

Thanks for the info. I see you are also still running v4.2.0 (4039).

I purposely waited till today to do more testing of how the following tags work (or don't).

So going back the #UpgradeAlarm did work yesterday it returned (1), though after a restart of CMX it reverted to (0) and I had still not upgraded.
Later last night I noticed it was (1) again (Mark how does this work).
I awoke this morning and it was (0), I restarted CMX and still (0) even 8 odd hours later. (note: still not upgraded yet either)

Mark you noted in the latest release v4.2.1 (b4042) "Fix web tag <#NewBuildNumber> showing the latest build number as "0000"".

Has #NewBuildAvailable been fixed as well because in my current v4.2.0 (4039) #NewBuildAvailable has never changed from (0) ever.

Both were reported as "Fix <#NewBuildAvailable> and <#NewBuildNumber> web tags" in v4.1.0 - (4024), though neither ever worked after that.

#version 4.2.0
#build 4039
#NewBuildAvailable 0
#NewBuildNumber 0000
#update 18/10/2024 2:15:10 PM
#UpgradeAlarm 0

5 min update of the tags here : https://beaumaris-weather.com/UpgradeChecks.php

Kindest Regards,

[mcrossley]

Here is what I get with b4041...

Code: Select all

{
    "build": "4041",
    "NewBuildAvailable": "1",
    "NewBuildNumber": "4042"
}

[BeaumarisWX]

Hi Mark,

Thanks for that : upgraded around 03:00 this morning (all worked fine ) Thanks.

Code: Select all

{
  "build": "4042",
  "NewBuildAvailable": "0",
  "NewBuildNumber": "4042"
}

Went to bed and you would not believe it (woke up at 06:30 to find CMX had shut down at 05:06) due to COM3 denied.

That's a new one for me checked and made sure power setting un-check on COM3 (all I can imagine is it may have been my UPS Recharging ?).

Obviously not a CMX cause, so will have to keep an eye on it.

Below is the later portion of the event, started around 05:06.

Code: Select all

2024-10-19 05:17:00.263 *** Data input appears to have stopped
2024-10-19 05:17:07.928 InitSerial: Connecting to the station
2024-10-19 05:17:07.966 InitSerial: Error opening port - Access to the path 'COM3' is denied.
2024-10-19 05:17:07.966 InitSerial: Failed to connect to the station, waiting 30 seconds before trying again
2024-10-19 05:17:37.971 InitSerial: Connecting to the station
2024-10-19 05:17:37.972 InitSerial: Error opening port - Access to the path 'COM3' is denied.
2024-10-19 05:17:37.972 InitSerial: Failed to connect to the station, waiting 30 seconds before trying again
2024-10-19 05:18:00.264 *** Data input appears to have stopped
2024-10-19 05:18:00.264 *** Exiting Cumulus due to Data Stopped condition for > 10 minutes
2024-10-19 05:18:06.515 Cumulus terminating
2024-10-19 05:18:06.516 Cumulus closing
2024-10-19 05:18:06.516 Stopping timers
2024-10-19 05:18:06.516 LOOP: Reconnected to station
2024-10-19 05:18:06.516 Stopping extra sensors...
2024-10-19 05:18:06.516 Ending normal reading loop
2024-10-19 05:18:06.516 Extra sensors stopped
2024-10-19 05:18:06.516 Stopping station...
2024-10-19 05:18:06.516 Writing today.ini file
2024-10-19 05:18:06.588 Completed writing today.ini file
2024-10-19 05:18:06.589 SaveWindData: Attempting to save the wind speeds array
2024-10-19 05:18:06.933 SaveWindData: Saved the wind speeds array
2024-10-19 05:18:06.934 Closing connection
2024-10-19 05:18:06.934 Station stopped
2024-10-19 05:18:06.934 Releasing lock file...
2024-10-19 05:18:07.245 Station shutdown complete

Kindest Regards,

[mcrossley]

That's an odd one.