The handling of large numbers of zone transfers between master and slave is not working well. Having several thousands of zones and a new slave DNS on the same internal network made me change the low defaults to this:
master: transfers-out 20; transfers-per-ns 20; slave: transfers-in 20; transfers-per-ns 20; It had catastrophic results. The slave transferred about 400 zones and then displaying a lot of "skipping zone transfer (cached)" messages and silence. Ordinary dig and rndc requests to both servers were working fine both with UDP and TCP also from the slave server. After 10 minutes of no action a repeat was performed. The same after another 10 minutes, then no action for many minutes. Changing to transfers-per-ns 10; made no difference. I had to go down to transfers-per-ns 4; before things were working. Something is not right here and should be looked into. - Jørgen Th _______________________________________________ bind-users mailing list bind-users@lists.isc.org https://lists.isc.org/mailman/listinfo/bind-users
