Vounerability if master dies - redundancy needed or not?

Ronni Jensen roj at mvb.dk
Thu Nov 23 09:19:51 UTC 2006


I have 2 issues I am a little uncertain about, so I'm gonna ask the
pro's here :) I have 1 hidden master, and 2 public slave servers.

The zone files SOA contains:

                                10800           ; Refresh
                                3600            ; Retry
                                604800          ; Expire
                                86400           ; Minimum

Question 1) If the master crashes or for some the slaves cannot reach
it, will the slaves keep serving their known data, or will their
zonefiles/cache be deleted when they cannot update from the master?

Question 2) In case that the answer to question 1 is that it will still
work regardless of how long the master is down, what will happen if I
reboot one of the slave servers? Will the slave still have the zonefiles
stored that it got transferred from the master earlier? Or will both
it's cache and zones be flushed when rebooting?

The point of these questions is to verify whether it is absolutely
critical that the master BIND also has redundancy like that slaves, or
it is ok that the master is down for X hours/days in e.g. case of a
restore from crash or hardware failure etc.

Thank you in advance :)

Best regards,

More information about the bind-users mailing list