Since a few days one of our devices in balena cloud is “Online (vpn only)”. Apparently that means the the supervisor is not able to do API calls.
There have been no (system) updates on the device as far as I known and a reboot does not help.
The supervisor logs are full of these messages:
[api] GET /v1/healthy 200 - 1.295 ms
[error] LogBackend: unexpected error: Error: read ECONNRESET
[error] at TLSWrap.onStreamRead (internal/stream_base_commons.js:205:27)
[error] LogBackend: unexpected error: Error: read ECONNRESET
[error] at TLSWrap.onStreamRead (internal/stream_base_commons.js:205:27)
any idea what could be causing this?
Hello, is the device acting normally otherwise? For example, have you successfully pushed a release to it since it started reporting this status?
hi @alanb128 ,
Thanks for checking in. Yes, everything seems to be acting normally, I can also push new/older releases to the device without any issues.
Does this mean that the supervisor is not able to send a heartbeat to the balena cloud?
As a long shot I tried to upgrade the supervisor from 12.11.30 to 14.10.10 but logs still report errors ( and the state is still ‘Device online (vpn only)’)
any ideas?
[debug] Finished applying target state
[success] Device state apply success
[info] Reported current state to the cloud
[error] LogBackend: unexpected error: Error: read ECONNRESET
[error] at TLSWrap.onStreamRead (node:internal/stream_base_commons:217:20)
[error] LogBackend: server responded with status code: 502
[error] LogBackend: unexpected error: Error: read ECONNRESET
[error] at TLSWrap.onStreamRead (node:internal/stream_base_commons:217:20)
[error] LogBackend: unexpected error: Error: read ECONNRESET
[error] at TLSWrap.onStreamRead (node:internal/stream_base_commons:217:20)
[api] GET /v1/healthy 200 - 3.897 ms
Hi, a balenaCloud API update was release a few hours ago to fix cached online status that are outdated. Can you confirm if your device still stuck with the Online (vpn only)
status?
Hi @cmfcruz ,
Yes, I can confirm it works now. I actually took the device home to rule out that any recent infrastructure changes on our network were the culprit. When I plugged it in 16:00 CET 2023-05-08 it worked, but I guess that was about the time that the fix rolled out?
regards,
Robert
Yes, right, looks like the times match. Since then, did you try to plug the device again into its original location (i.e., not your home)? Did it went “fully online” normally?
yep. I plugged it back in at work the next day and it reported to be “fully” online directly.