dhcpd peer constantly requesting rebalancing

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

dhcpd peer constantly requesting rebalancing

Brian De Wolf
Hello,

We have two servers running dhcpd-4.3.2 as failover peers.
Every other week or so we'll see them get stuck requesting rebalancing
for a subnet every minute.  This goes on for a day or two until it
somehow resolves itself (I assume some leases expire).  It has occurred
to different subnets, so it doesn't seem like an issue from a specific
area.

Has anyone seen this issue?  Does it represent a problem?  Is there a
way to resolve it manually?

I read through the release notes for 4.3.3 but didn't see any relevant
changes.  The closest was related to delayed-ack, which we do not have
enabled.
_______________________________________________
dhcp-users mailing list
[hidden email]
https://lists.isc.org/mailman/listinfo/dhcp-users
Reply | Threaded
Open this post in threaded view
|

Re: dhcpd peer constantly requesting rebalancing

Simon Hobson
Brian De Wolf <[hidden email]> wrote:

> Every other week or so we'll see them get stuck requesting rebalancing
> for a subnet every minute.  This goes on for a day or two until it
> somehow resolves itself (I assume some leases expire).

How many free leases are there when this happens ?

_______________________________________________
dhcp-users mailing list
[hidden email]
https://lists.isc.org/mailman/listinfo/dhcp-users
Reply | Threaded
Open this post in threaded view
|

Re: dhcpd peer constantly requesting rebalancing

Brian De Wolf
On Wed, 16 Mar 2016 10:09:27 +0000
Simon Hobson <[hidden email]> wrote:

> Brian De Wolf <[hidden email]> wrote:
>
> > Every other week or so we'll see them get stuck requesting
> > rebalancing for a subnet every minute.  This goes on for a day or
> > two until it somehow resolves itself (I assume some leases
> > expire).  
>
> How many free leases are there when this happens ?
>

It seems to vary.  One example:

        server 1: balancing pool 24ea320 134.71.22.0/26  total 51  free 1  backup 7  lts -3  max-own (+/-)1  (requesting peer rebalance!)
        server 2: balancing pool 2a2e300 134.71.22.0/26  total 51  free 3  backup 5  lts 1  max-own (+/-)1

Another:

        server 1: balancing pool 1c06510 134.71.200.0/24  total 249  free 68  backup 48  lts 10  max-own (+/-)12
        server 2: balancing pool 214a4f0 134.71.200.0/24  total 249  free 93  backup 23  lts -35  max-own (+/-)12  (requesting peer rebalance!)

Sometimes the subnet is near full, sometimes it isn't.
_______________________________________________
dhcp-users mailing list
[hidden email]
https://lists.isc.org/mailman/listinfo/dhcp-users