Created attachment 14268 [details]
Error log 1
When I try to update DNS entries from my RODC, I got error (in my_rodc_error.log attachment).
Moreover, my client tried to do the same and he got another error (client_rodc_error.log).
That's strange because if you manually add the DNS entry, it works well.
Created attachment 14269 [details]
Error log 2
Created attachment 14526 [details]
Succession of calls to samba_dnsupdate, showing non-updates entries
I am working on this issue, and found the following :
- Default-First-Site-Name (associated to subnet 192.168.56.0/24)
- Secondary (associated to subnet 192.168.57.0/24)
- sambarwdc - DC - on site "Default-First-Site-Name"
- sambarodc - RODC - on site "Secondary"
After joining the domain as an RODC, when I run the "samba_dnsupdate" command, the RODC makes the "failed DNS update" errors reported above, associated to a timeout error.
But if you check those entries (using "host -t SRV [entry]"), they will actually be found.
Yet if you call "samba_dnsupdate" again, there are two entries that keep being updated - and this is repeatable each time you run the command.
Testing those two entries show that they are effectively not registered in the DNS.
The log attached shows the following sequence:
- samba_dnsupdate right after joining as an RODC
- several new calls to show the same entries being updated
- verifying the entries are not found using "host -t SRV"
I feel that the initial failure due to Timeout are actually not a bug - maybe the timeout is too short, but apparently the update succeeds in the end, so it's not really an issue.
Now for the two entries that keeps being updated - even though no error are shown when we update them - this is a concern.
Why are those updates failing? Is there something we can do to fix that, or is it a bug?
Any advice is welcome to troubleshoot this.