Re: DHCP Clustering on Windows 2003 - HELP!!!
- From: "Reginald Francois" <rfrancoi@xxxxxxxxxxxx>
- Date: Tue, 7 Jun 2005 19:30:34 -0400
Ok Charles. I understand most of what you are telling me.
Here is another scenerio:
I have 2 machines, each having a SCSI controller, and 1 HDD. First of
all, is it possible for me to create a cluster?
On one machine, at one point I had 2 SCSI HDDs. The Quorum was created on
one of them, and the on the other, the "F" drive appeared. When I
disconnected the second drive on that machine and started from scratch
(Ghosted Win2K3SP1 image), I was not able to have the second machine join
the Cluster.
Is this because the first node took the first SCSI (ID:0) as the Quorum and
the second SCSI (ID:1) as the shared disk?
Thanks,
Reggie
"Charles Tolento" <CharlesTolento@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in
message news:D48DA609-CB30-41D3-B97D-A97E4FE2E2D9@xxxxxxxxxxxxxxxx
> 1. Reginald
>
> 1. IIS Common Files are needed to install cluster services
>
> 2. Yes shut down the node so both nodes do not attempt to
> access the shared stoage. Once a node has cluster services install it
> will clusdsk.sys will protect the cluster disks. The message is normal
> because you are probably using a crossover cable for the heartbeat NIC.
> Just
> make sure you configure the Heartbeat network once both nodes are
> installed
> via KB 258750.
>
> 3. Resources in a cluster are only available to one node at any one time.
> Your Disk is seen by the first node because you installed Cluster SErvices
> there first. If you fail over the disk resource group NODE 2 would then
> see
> the disk.
>
> 4. DHCP resource requires IP, Physical Disk and Network Name resource
> dependencies. You should not use the default Cluster Name or Quorum Disk
> for
> this purpose. http://support.microsoft.com/?kbid=226796
>
> 5. By defaul a resource has to fail more than 3 times in 15 minutes in
> order for failover to occurr. If you initiate failover 4 times in a row
> you
> will see the resource group failover. You are able to see the DHCP scope
> on
> either node because your are connecting to the Virtual Server name. The
> Vitrual Server name is online so through AD and WMI you are able to
> connect
> to any DHCP server in AD via the DHCP snap in.
>
> 6. If you want to see which network is being used for Heartbeat
> communication you can use Network Monitor and look for UDP packets on port
> 3343. Only one network will send the heartbeat traffic at a time.
>
> 7. DHCP Link for w2k is provided above the process is the same for w2k3.
>
> Regards
>
> CT
>
> For more cluster administration tips and tools please visit my website
> that
> is currently being developed in the future.
>
> www.cluadmin.com
>
>
>
> "Reginald Francois" wrote:
>
>> Hello everyone,
>>
>> I have a few more questions regarding DHCP clustering that I hope
>> someone
>> can help me with.
>>
>> 1. Is IIS required for clustering? During the configuration of the
>> first
>> node on my cluster, I was prompted for the Windows 2003 CD.
>>
>> 2. Should the machine that is going to be used as the second node be
>> completely turned off during the intial Cluster configuration? When I
>> did
>> that, I observed a messages referring the private NIC on the secondary
>> not
>> being detected. In the
>> documentation that I've read, the instructions are to turn off all
>> machine
>> that will be used as nodes on the cluster off to prevent corrupting the
>> DataBase. Are they referring to the
>> Cluster DataBase or the resource DataBase(DHCP). At this point, there
>> is
>> no DHCP DataBase to corrupt, am I correct? Should I be leaving the
>> secondary machine on
>> duing the creation of the Cluster?
>>
>> 3. After the configuration of the first node on my cluster, a new
>> drive
>> appeared ("F:") on the first node, but not on the second. Is this
>> expected?
>> Is this the
>> drive that will be used for my FailOver? I noticed that the DHCP
>> DataBase
>> is on this drive.
>>
>> 4. Is the creation of a Virtual Server required for a DHCP cluster,
>> or
>> is it recreated automatically? When I try to create on, and select the
>> DHCP
>> Service as a
>> resource, I get a error message referring to "Physical Disk". What
>> should I
>> do?
>>
>> 5. When I do a "Initiate Fail Over", the resources do not move to my
>> secondary node. Although, when I launch DHCP on my second node, all of
>> the
>> scope options and lease information are present.
>>
>> 6. Fail Over test: In the documentation that I've read, it is
>> recommended that the "Private Network" NIC be configured for "Node-Node"
>> communication only. And the "Public Network" be configured for ALL
>> communication. How can I test if the primary node is going to be
>> detected
>> on the "Public Network" NIC? I
>> want to be able to test the complete transfer of rolls to the secondary
>> Node. How do I do this this? Which NIC is used to transfer the DHCP
>> DataBase to the secondary?
>>
>> 7. Can someone point me to some solid (informative) reference
>> material
>> on clustering with DHCP on Windows 2003? Any book or website would be
>> GREATLY appriciated.
>>
>> Thank you in advance,
>> Reginald Francois
>> rxf1@xxxxxxx
>>
>>
>>
.
- References:
- DHCP Clustering on Windows 2003 - HELP!!!
- From: Reginald Francois
- RE: DHCP Clustering on Windows 2003 - HELP!!!
- From: Charles Tolento
- DHCP Clustering on Windows 2003 - HELP!!!
- Prev by Date: Re: MSDTC failure after Windows Server 2003 Service Pack 1 Install
- Next by Date: A scenerio - 2 Machines, 1 IDE, 1 SCSI drive each
- Previous by thread: RE: DHCP Clustering on Windows 2003 - HELP!!!
- Next by thread: Cannot start Cluster Resources
- Index(es):
Relevant Pages
|
Loading
