Re: IP Address resource does not come up after failover!
- From: "Chuck Timon [Microsoft]" <ctimon@xxxxxxxxxxxxxxxxxxxx>
- Date: Mon, 14 Aug 2006 08:10:50 -0400
That's not the way cluster works. Cluster does not assign any IP
addresses - it brings them online provided the network supporting the IP
address is up. Cluster is just a multi-homed machine. The process cluster
uses is here - http://support.microsoft.com/default.aspx?scid=kb;[LN];168567
Please also tell me that since these NICS are plugged into different
switches, they are on the same VLAN? Seperate VLAN? What?
--
Chuck Timon, Jr.
Microsoft Corporation
Longhorn Readiness Team
This posting is provided "AS IS" with no
warranties, and confers no rights.
"Max2131" <Max2131@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:792A4048-EC12-4D1F-9B74-45F504C4AF9F@xxxxxxxxxxxxxxxx
Hi Chuk. Thanks for the reply.
The problems is that the network is not down in a sense that there is a
good
connection between the NICs of both machines and the switches (links are
up).
The failure is somewhere in the middle of the network which affects the
connectivity between the cluster nodes.
However, if I, say, just unplug one of the NICs from its switch the
cluster
service will correctly do the failover to the node that still has the NIC
connected to the switch.
So here is the problem - why during a manual failover to a cluster node
does
not the cluster service just assign the resource IP address to the node's
NIC
(which is connected to the switch and its link is up) without proper
connection to the cluster's other node when its NIC also properly
connected
to the switch?
"Chuck Timon [Microsoft]" wrote:
If the network that an IP address resource is associated with is down,
the
IP address will fail - expected behavior. Because the network using the
crossover cable is still up the cluster service on both nodes continues
to
run because it can communicate with all the nodes over that network. One
of
those communications is the fact that there is a failed IP address in the
cluster.
--
Chuck Timon, Jr.
Microsoft Corporation
Longhorn Readiness Team
This posting is provided "AS IS" with no
warranties, and confers no rights.
"Max2131" <Max2131@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:282C5130-5445-422B-9C3B-A3BF5C2A7F25@xxxxxxxxxxxxxxxx
Hi. I use a two machine cluster (Server 2003 R2 EE). Each of them has
two
Ethernet NICs and both are connected to a SCSI quorum.
There are two Ethernet paths between the machines - one is just a
crossover
(never fails ;) and the other is a more complicated connection via
several
switches.
Each of these network connections has an IP address resource assigned
to,
and both of these resources are in the same group as the quorum.
Here is the problem:
if, for some reason, one of the intermediate switches goes off (so the
second connection gets broken) and I do a manual failover (which
otherwise
works just fine) - the IP address resource for the broken connection
never
comes up, and the group marked as failed. I expected that, even in the
case
of failure, given that there still is a working connection between the
machines (crossover) this IP address resource should just move to the
designated host...but it does not work.
Does anybody have an idea?
I'd really appreciate any help you guys can provide.
.
- Follow-Ups:
- Re: IP Address resource does not come up after failover!
- From: Max2131
- Re: IP Address resource does not come up after failover!
- References:
- Re: IP Address resource does not come up after failover!
- From: Chuck Timon [Microsoft]
- Re: IP Address resource does not come up after failover!
- Prev by Date: Re: help NEWBIE!
- Next by Date: Re: eSATA Cluster Support?
- Previous by thread: Re: IP Address resource does not come up after failover!
- Next by thread: Re: IP Address resource does not come up after failover!
- Index(es):
Relevant Pages
|
