Dell Networking Force10 switches and Microsoft Network Load Balancing


Dell Networking Force10 switches and Microsoft Network Load Balancing


This article explains and brief overview of Microsoft Network Load Balancing (NLB) and how to set that up on Dell Networking Force10 switches.



Objectives
  1. NLB Overview
  2. NLB Unicast mode
  3. NLB Multicast mode
  4. Limitations with enabling NLB on switches
  5. Benefits working with Microsoft Clustering
  6. Enable and Disable VLAN flooding
  7. Configuring a switch for NLB


NLB Overview

This functionality is supported on Dell Networking OS.
Network Load Balancing (NLB) is a clustering functionality that is implemented by Microsoft on Windows 2000 Server and Windows Server 2003 operating systems. NLB uses a distributed methodology or pattern to equally split and balance the network traffic load across a set of servers that are part of the cluster or group. NLB combines the servers into a single multicast group and attempts to use the standard multicast IP or unicast IP addresses, and MAC addresses for the transmission of network traffic. At the same time, it also uses a single virtual IP address for all clients as the destination IP address, which enables servers to join the same multicast group in a way that is transparent to the clients (the clients do not notice the addition of new servers to the group). The clients use a cluster IP address to connect to the server. The NLB functionality enables flooding of traffic over the VLAN ports (for unicast mode) or a subset of ports in a VLAN (for multicast mode) to avoid overloading and effective performance of the servers for optimal processing of data packets.
NLB functions in two modes, namely unicast mode and multicast mode. The cluster IP address and the associated cluster MAC address are configured in the NLB application running on the Windows Server. In the unicast mode, when the server IP address is attempted to be resolved to the MAC address using the ARP application, the switch determines whether the ARP reply, obtained from the server, is of an NLB type. The switch then maps the IP address (cluster IP) with the MAC address (cluster MAC address). In multicast mode, the cluster IP address is mapped to a cluster multicast MAC address that is configured using a static ARP CLI configuration command. After the NLB entry is learned, the traffic is forwarded to all the servers in the VLAN corresponding to the cluster virtual IP address.

Microsoft Network Load Balancing feature is only available on in FTOS versions 9.3.0.0 and higher.

NLB Unicast Mode
Consider a sample topology in which four servers, namely S1 through S4, are configured as a cluster or a farm. This set of servers is connected to a Layer 3 switch, which in turn is connected to the end-clients. The servers contain a single IP address (IP-cluster address of 172.16.2.20) and a single unicast MAC address (MAC-Cluster address of 00-bf-ac-10-00-01) for load-balancing. Because multiple ports of a switch cannot learn a single MAC address, the servers are assigned with MAC addresseses of MAC-s1 to MAC-s4) respectively on S1 through S4 in addition to the MAC cluster address. All the servers of the cluster belong to the VLAN named VLAN1.
In unicast NLB mode, the following sequence of events occurs:
  • The switch sends an ARP request to resolve the IP address of the cluster MAC address.
  • The ARP servers send an ARP response with the MAC cluster address in the ARP header and a MAC address of MAC-s1/s2/s3/s4 (for servers S1 through S4) in the Ethernet header.
  • The switch associates the IP address with the MAC cluster address with the last ARP response it obtains. Assume that in this case, the last ARP reply is obtained from tMAC-s4 (assuming that the ARP response with MAC-s4 is received as the last one). The interface associated with the server, S4, is added to the ARP table.
  • With NLB feature enabled, after learning the NLB ARP entry, all the subsequent traffic is flodded on all ports in VLAN 1.

With NLB, the data frame is forwarded to all the servers for them to perform load-balancing.

NLB Multicast Mode

Consider a sample topology in which four servers, namely S1 through S4, are configured as a cluster or a farm. This set of servers is connected to a Layer 3 switch, which in turn is connected to the end-clients. They contain a single multicast MAC address (MAC-Cluster: 03-00-5E-11-11-11).
In the multicast NLB mode, a static ARP configuration command is configured to associate the cluster IP address with a multicast cluster MAC address.
With multicast NLB mode, the data is forwarded to all the servers based on the port specified using the Layer 2 multicast command, which is the mac-address-table static <multicast_mac> multicast vlan <vlan_id> output-range <port1>, <port2> command in CONFIGURATION mode.



Limitations with Enabling NLB on switches
  • ​The NLB unicast mode uses switch flooding to transmit all packets to all the servers that are part of the VLAN. When a large volume of traffic is processed, the clustering performance might be impacted in a small way. This limitation is applicable to switches that perform unicast flooding in the software.
  • The ip vlan-flooding command applies globally across the system and for all VLANs. In cases where the NLB is applicable and the ARP replies contain a discrepancy in the Ethernet SHA and ARP header SHA frames, a flooding of packets over the relevant VLAN occurs.
  • The maximum number of concurrent clusters that is supported is eight.
Benefits working with Microsoft Clustering

Microsoft clustering allows multiple servers using Microsoft Windows to be represented by one MAC address and IP address in order to provide transparent failover or balancing. Dell Networking OS does not recognize server clusters by default; it must be configured to do so. When an ARP request is sent to a server cluster, either the active server or all the servers send a reply, depending on the cluster configuration. If the active server sends a reply, the Dell switch learns the active server’s MAC address. If all servers reply, the switch registers only the last received ARP reply, and the switch learns one server’s actual MAC address; the virtual MAC address is never learned. Because the virtual MAC address is never learned, traffic is forwarded to only one server rather than the entire cluster, and failover and balancing are not preserved.
To preserve failover and balancing, the switch forwards the traffic destined for the server cluster to all member ports in the VLAN connected to the cluster. To ensure that this happens, you must configure the ip vlan-flooding command on the Dell switch at the time that the Microsoft cluster is configured. The server MAC address is given in the Ethernet frame header of the ARP reply, while the virtual MAC address representing the cluster is given in the payload. Then, all the traffic destined for the cluster is flooded out of all member ports. Since all the servers in the cluster receive traffic, failover and balancing are preserved


Enable and Disable VLAN Flooding
  • The older ARP entries are overwritten whenever newer NLB entries are learned.
  • All ARP entries, learned after the feature is enabled, are deleted when the feature is disabled, and RP2 triggers an ARP resolution. The feature is disabled with the no ip vlan-flooding command.
  • When a port is added to the VLAN, the port automatically receives traffic if the feature is enabled. Old ARP entries are not deleted or updated.
  • When a member port is deleted, its ARP entries are also deleted from the CAM.
  • Port channels in the VLAN also receive traffic.
  • There is no impact on the configuration from saving the configuration.
  • The feature, if enabled, is displayed in the show running-config command output that displays the ip vlan-flooding CLI configuration. Apart from it, there is no indication of the enabling of this capability.


Configuring a Switch for NLB

To enable a switch for Unicast NLB mode of functioning, perform the following steps:
.
Command Parameters
FTOS# configuration Enter global configuration mode.
FTOS(conf)# ip vlan-flooding This specifies that all Layer 3 unicast routed data traffic, going through a VLAN member port, needs to be flooded across all the member ports of that VLAN. There might be some ARP table entries that are resolved through ARP packets, which had the Ethernet MAC SA different from the MAC information inside the ARP packet. This unicast data traffic flooding occurs only for those packets that use these ARP entries.


To enable a switch for Multicast NLB mode of functioning, perform the following steps:

Command Parameters
FTOS# configure Enter global configuration mode.
FTOS(conf)# arp 192.168.1.1 <multicast MAC address> te 0/10 This command adds a static ARP entry to associate an IP address with a multicast MAC address in the switch. This setting causes the multicast MAC address to be mapped to the cluster IP address for the NLB mode of operation of the switch.
FTOS(conf)# mac-address-table static <multicast MAC address> multicast vlan 1 output- range te 0/1 - 4 Associate specific MAC or hardware addresses to VLANs







Article ID: HOW10789

Last Date Modified: 09/14/2019 12:27 AM

Rate this article

Accurate
Useful
Easy to understand
Was this article helpful?
0/3000 characters
Please provide ratings (1-5 stars).
Please provide ratings (1-5 stars).
Please provide ratings (1-5 stars).
Please select whether the article was helpful or not.
Comments cannot contain these special characters: <>()\
characters left.