Solved

NLB with 2 NIC's

Posted on 2006-06-13
9
765 Views
Last Modified: 2012-06-22
I am attempting to setup NLB with 2 network cards (as recommended) on my LAN, and would like to know if this will cause routing problems in the following scenario.

Cluster IP 192.168.1.100 > all /24 - all NIC's plugged into the same Netgear router/switch

**************************
Server1

NIC #1 (cluster host IP)
IP                 192.168.1.101
Gateway        192.168.1.1
DNS              192.168.1.2

NIC #2 (other LAN traffic IP)
IP                192.168.1.3
Gateway        192.168.1.1
DNS              192.168.1.2      

**************************
Server2

NIC #1 (cluster host IP)
IP                 192.168.1.102
Gateway        192.168.1.1
DNS              192.168.1.2

NIC #2 (other LAN traffic IP)
IP                192.168.1.4
Gateway        192.168.1.1
DNS              192.168.1.2  

So far, I am having routing problems on Server 1, but not Server 2, and just need to know if this is a problem having all NIC's pointing to the same gateway or if this is an acceptible scenario.  Something tells me that this is not correct, but I thought I would get confirmation from an expert.  Thanks.
0
Comment
Question by:bleujaegel
  • 5
  • 3
9 Comments
 
LVL 33

Accepted Solution

by:
NJComputerNetworks earned 325 total points
ID: 16901983
My setup looks more like this  

**************************
Server1

NIC #1 (cluster host IP)
IP                 192.168.1.101
Gateway        192.168.1.1
DNS              192.168.1.2

NIC #2 (Internal Heartbeat communication)  (Connected to other server via cross over cable or hub)
IP                10.10.10.1
Gateway        
DNS              (Host File with Servername2   10.10.10.2)      

**************************
Server2

NIC #1 (cluster host IP)
IP                 192.168.1.102
Gateway        192.168.1.1
DNS              192.168.1.2

NIC #2 (Internal Heartbeat communication)  (Connected to other server via cross over cable or hub)
IP                10.10.10.2
Gateway        
DNS              (Host File with Servernam1   10.10.10.1)  
0
 
LVL 33

Expert Comment

by:NJComputerNetworks
ID: 16902011
0
 
LVL 33

Expert Comment

by:NJComputerNetworks
ID: 16902049
0
NEW Veeam Agent for Microsoft Windows

Backup and recover physical and cloud-based servers and workstations, as well as endpoint devices that belong to remote users. Avoid downtime and data loss quickly and easily for Windows-based physical or public cloud-based workloads!

 
LVL 2

Author Comment

by:bleujaegel
ID: 16904123
Ok, I see.  Keep the NLB's connected directly via a separate subnet, using a host file for 'DNS' resolution.  Why are the default gateways left blank?  Does the OS automatically add the necessary routes to the routing table to allow for inbound/outbound communications?

Would it be correct to say that the 10.0.0.0 subnet is strictly for communications between the two servers, and no outside traffic?  I'm guessing that you are running unicast mode.  Also, are the external requests are routed to the NIC's with the .101 & .102 IP's only?  From what I've read, it sounds like you have 2 NIC's, the 192.168.1.101 & 102 running multicast mode, and the 10.10.10.1 & 2 running unicast.  Am I close?  Thanks.
0
 
LVL 2

Author Comment

by:bleujaegel
ID: 16906600
I'm assuming I could use a switch instead of a crossover cable for heartbeat communications, which would then allow me to scale out at multiple NLB servers.  Would this be correct?
0
 
LVL 33

Expert Comment

by:NJComputerNetworks
ID: 16911172
"Why are the default gateways left blank?"  Blank because no routing is needed... crossover cable doesn't need a gateway..  If you place a switch here, you want' need a default gateway either unless you have multiple networks.

"Would it be correct to say that the 10.0.0.0 subnet is strictly for communications between the two servers, and no outside traffic? "   Yes, this is correct...

This is just the way I set it up...  I don't know if it is optimal...but it seems to work for my environment...  
0
 
LVL 84

Assisted Solution

by:oBdA
oBdA earned 175 total points
ID: 16912192
Sorry, but either you're talking apples and oranges here, or the NLB setup is incorrect (though working).
Only a *server* *cluster* requires a separate heartbeat network.
Network Load Balancing does NOT require a separate heartbeat, as the NLB heartbeat will be sent (only!) over the *clustered* network cards. Check here:
Network Load Balancing Frequently Asked Questions
http://www.microsoft.com/technet/prodtechnol/windowsserver2003/technologies/clustering/nlbfaq.mspx

Now to your setup:
Whether you require one or two NICs is mainly dependant on the mode you're running your NLB cluster in, and whether you need inter-host communication.
Multicast requires only one NIC even if you need inter-host communication; this will work well if there is only one subnet to be considered. If you have more than one subnet, you'll have to add static ARP entries on your router(s), as (most) routers won't add the multicast MAC address.
Unicast requires two NICs if you need inter-host communication. The advantage is that an NLB cluster in unicast mode can be reached from other subnets without problems.
So your inital setup should basically work; just remove the default gateway from the non-clustered NICs, and use unicast mode (or leave it in multicast, and use only one NIC while adding static ARP entries to the routers as I described above).
Here are some additional links that might help:

Network Load Balancing: Configuration Best Practices for Windows 2000 and Windows Server 2003
http://www.microsoft.com/technet/prodtechnol/windowsserver2003/technologies/clustering/nlbbp.mspx

Using Clustering for a Highly Available Web Site: An Example
http://www.microsoft.com/technet/prodtechnol/windowsserver2003/technologies/clustering/scenep2.mspx

Network Load Balancing clusters
http://www.microsoft.com/resources/documentation/WindowsServ/2003/standard/proddocs/en-us/microsoft_WLBS.asp

Best practices
http://www.microsoft.com/resources/documentation/WindowsServ/2003/standard/proddocs/en-us/NLB_best_practices.asp
0
 
LVL 33

Expert Comment

by:NJComputerNetworks
ID: 16912345
Your correct OBdA,  I must have added the second NIC (not for a heartbeat) but because I found my hosts could not communicate with one another:

Inter-host communication in unicast mode

In unicast mode, each host in the cluster has the same IP Address and the same MAC Address making them look identical from a networking perspective. So, unicast mode has the side effect of disabling communication among the hosts of the cluster.

0
 
LVL 2

Author Comment

by:bleujaegel
ID: 17006467
Working good.  I will try unicast mode next...

Thanks for the help.
0

Featured Post

Free Tool: Path Explorer

An intuitive utility to help find the CSS path to UI elements on a webpage. These paths are used frequently in a variety of front-end development and QA automation tasks.

One of a set of tools we're offering as a way of saying thank you for being a part of the community.

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

Organizations create, modify, and maintain huge amounts of data to help their businesses earn money and generally function.  Typically every network user within an organization has a bit of disk space to store in process items and personal files.   …
While rebooting windows server 2003 server , it's showing "active directory rebuilding indices please wait" at startup. It took a little while for this process to complete and once we logged on not all the services were started so another reboot is …

680 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question