-
Notifications
You must be signed in to change notification settings - Fork 243
Description
If the device finishes initialising BEFORE the network comes up, NTP fails and, as far as I can tell, is never retried.
It seems to go something like -
- device powers on
- jetkvm_app starts
- jetkvm_app requests NTP and connects to cloud (all fails)
- network connects, gets IP.
I can SSH in, and see date = 1970. Cloud connection not present.
$ ssh root@192.168.11.61 date
Thu Jan 1 00:01:10 UTC 1970
You can simulate this just by plugging the power in, waiting for it to finish booting, then connecting the network.
Once it has failed to obtain NTP, all SSL certificates are considered invalid, so it wont sign into the cloud service, will
return certificate failures when checking for updated.
As far as I can tell, it never retries the NTP. I left it running for a day, and the internal clock never synced.
This appears to be a aggravated by my switch taking excessively long to negotiate 100MB Fast Ethernet..
When I move the ethernet connection to a different switch, the network initialises faster, and everything is fine.
This manifests in many ways, including;
- Update Error - Invalid TLS certificate #534
- Time Synchronization does not work #538
- Error downloading 0.39 Update #350
- failed to check for updates - found workaround #68
- Time Synchronization does not work #538
- [Alternative Fix] Set system time without NTP #293
The root cause is always NTP failing, and not being retried.
I suggest that NTP should be retried rapidly until it has been set at least once, then return to its normal schedule.
Thanks