I frequently have 625 and 626 errors on certain remote servers when checking timesync due to our infrastructure being comprised of many, many, small satellite offices, many of which have less than optimal WAN links.
But this really takes the cake.
Server dhcplakehurst returns a 626 error during an ndsrepair -T operation.

However: not only does it resolve the IP and ping just fine by name, I also ran nmap against the tcp NCP port:
nmap dhcplakehurst -sS -p 524
and it came immediately back showing the port as "open".. no issue there at all.

I tried to do a repair on the network address using ndsrepair -E, and it fails too with a -626 error.

I don't get it.. what does it want?? I don't see what could be causing the problem.
I don't have a DNS issue, I don't have a network blockage issue. The pings returned in about 12ms, no problem or delay there. The WAN appears healthy enough.
Probably later today it will resolve itself, as it often does, only to return again soon.. it's quite possible that tomorrow morning or the day after, it will do this again.