[nsd-users] Interpreting log messages during inbound XFR

[Klaus Darilion]

Hi!

I just analyzed why an IXFR takes longer on one server than on another:

Server 1
14:21:03 vie1 nsd[25852]: notify for xxx. from 11.22.33.44 serial 2019011730
14:21:04 vie1 nsd[669]: xfrd: zone xxx committed "received update to
serial 2019011730 at 2019-01-17T14:21:04 from 11.22.33.44 TSIG verified
with key rcode0-distribution"
14:21:10 vie1 nsd[672]: zone xxx. received update to serial 2019011730
at 2019-01-17T14:21:04 from 11.22.33.44 TSIG verified with key
rcode0-distribution of 12522393 bytes in 0.448595 seconds
14:21:13 vie1 nsd[669]: zone xxx serial 2019011729 is updated to 2019011730.

Server 2
14:21:02 fra1 nsd[20402]: notify for xxx. from 11.22.33.44 serial 2019011730
14:21:03 fra1 nsd[22465]: xfrd: zone xxx committed "received update to
serial 2019011730 at 2019-01-17T14:21:03 from 11.22.33.44 TSIG verified
with key rcode0-distribution"
14:21:40 fra1 nsd[22466]: zone xxx. received update to serial 2019011730
at 2019-01-17T14:21:03 from 11.22.33.44 TSIG verified with key
rcode0-distribution of 12522393 bytes in 0.439546 seconds
14:21:44 fra1 nsd[22465]: zone xxx serial 2019011729 is updated to
2019011730.


What exactly took longer on server 2? The IXFR itself? Forking the
process which loads the new zone? Loading/hashing the new zone?

Thanks
Klaus