Failover strangeness

[Greg G]

Glenn Satchell wrote:
>>   The secondary dhcp server is down, but the running server is still 
>> showing as being in communications-interrupted.  It never gets to 
>> partner-down.
>>     
>
> Ahh, didn't realise the second server was down. partner-down mode must
> be set manually. There is one method described in the dhcpd.conf man
> page, by editting the leases file.
>
>   
   After much futher testing, it's still giving me heartburn.  Here's 
the situation that I'm finding:

1.  Primary handles a lease
2.  Primary goes down.  Secondard goes to communications-interrupted
3.  Secondary dhcp server handles a lease that would normally belong to 
the primary
4.  Primary comes back up.  Secondary goes to normal mode.
5.  Secondary continues to handle the lease.
6.  I told the primary to operate in partner-down.  It went back to 
normal mode in about 2 seconds.  (This also forces a peer update.)
7.  Secondary continues to handle the lease.
8.  The secondary goes down.  (Primary goes to communications-interrupted.)
9.  The primary complains that it has No Free Leases.
10.  I told the primary to operate in partner-down.
11.  The primary continues to complain that it has No Free Leases.
12.  About a minute later, Primary handles the lease.

   This is quite problematic, as it seems that any time the secondary 
handles a lease, the primary is unable to handle it again unless the 
secondary goes down AND is told to operate in partner-down.  That's bad, 
IMO.  The manual intervention needed because a lease was handled by the 
secondary some time in the past is likely to cause problems.  I had 
hoped that the step 6 forcing of partner-down would fix this, but it didn't.

   Note that I'm using an omshell to set local-state = 00:00:00:04 
(that's partner-down).

   Thanks.

-Greg G