Load Balancing Ansible Tower Using NSX

Disclamer: this configuration is not validated by either VMware or Red Hat. Make sure it is applicable to your use case and thoroughly test before implementing in production.

Overview

If you landed on this page I trust you already know what Ansible is. It’s a great configuration management tool centred around using YAML to describe the desired state configuration of your various infrastructure components. This desired state is captured in what Ansible calls playbooks, which once written, can then be used in a repeatable way to deploy brand new components or enforce configuration on already deployed ones.

Ansible can be installed and used from CLI, which is usually a good starting point. If you have multiple people using Ansible in your organization, you can also deploy AWX. It’s a free GUI add-on to Ansible, which makes managing concurrent user access to Ansible easier, by adding projects, schedules and credentials management. On top of that there is Ansible Tower. Ansible Tower is a paid version of AWX and gives you additional enterprise features and services like clustering, product support, validated upgrade paths, etc. In this article we will be focusing on Ansible Tower version of the product.

Also worth mentioning that this configuration will be based on Ansible Tower cluster feature, which lets you run all nodes as active/active. Prior to version 3.1 it was called redundancy and worked only in active/passive mode. Redundancy feature is deprecated and is outside the scope of this blog post.

Topology

Deploying multiple Ansible Tower nodes in a cluster already gives you redundancy. If one of the nodes fails you can connect to another node, by just changing your browser URL. The benefit of having a load balancer is that you have one URL you can hit and if a node goes down, such situation is handled by load balancer automatically.

In this example we will be deploying a VMware NSX load-balancer in the following topology:

Configuration

Deploying an NSX load-balancer for HTTPS port 443 is simple, you can find numerous examples of how to create application profiles, monitors, pools and VIPs in official VMware documentation or out on the Internet. But with Ansible there’s one catch. If you try to use the default HTTPS monitor that NSX load balancer comes with, you will find HTTP 400 code in Ansible nginx logs:

10.20.30.40 - - [20/Jan/2020:04:50:19 +0000] "GET / HTTP/1.0" 400 3786 "-" "-" "-"
10.20.30.40 - - [20/Jan/2020:04:50:24 +0000] "GET / HTTP/1.0" 400 3786 "-" "-" "-"
10.20.30.40 - - [20/Jan/2020:04:50:29 +0000] "GET / HTTP/1.0" 400 3786 "-" "-" "-"

And an error in NSX load balancer health check:

As it turns out, when you make a HTTP request to Ansible Tower, specifying HTTP “Host” header is a requirement. Host header simply contains the hostname of the server you’re making a request to. Browsers add this header automatically, that’s why you’re not going to see any errors, when accessing Ansible Tower Using Firefox or Chrome. But NSX doesn’t add this header to the monitor checks by default, which makes Ansible Tower upset.

Here is the trick you need to do to make Tower happy:

Now nginx logs show success code 200:

10.20.30.40 - - [21/Jan/2020:22:54:42 +0000] "GET / HTTP/1.0" 200 11337 "-" "-" "-"
10.20.30.40 - - [21/Jan/2020:22:54:47 +0000] "GET / HTTP/1.0" 200 11337 "-" "-" "-"
10.20.30.40 - - [21/Jan/2020:22:54:52 +0000] "GET / HTTP/1.0" 200 11337 "-" "-" "-"

Load balancer health check is successful:

And pool members are up and reachable:

Note: technically the host header should contain the hostname of the Tower node we’re making a health check on. But since NSX monitor is configured per pool and not per pool member, we have to use a fake hostname “any.host.com” as a workaround. When I was testing it, Tower didn’t complain.

Reference

Even though I said that the rest of the load-balancer configuration is standard, I still think having screenshots for reference is helpful if you need to validate configuration. So find the full list of settings below.

Screenshot 1: Application Profile

Screenshot 2: Service Monitor

Screenshot 3: Pool

Screenshot 4: Virtual Server

Advertisement

Reminder: Disable Firewall on NSX ECMP Edge

ECMP and Stateful Services

It’s not new, this topic has already been discussed many times before, examples are here, here, here and here. When NSX Edges are configured in ECMP mode, none of the stateful services like VPN, NAT or Load Balancing are supported.

From NSX Design Guide:

In ECMP mode, only routing service is available. Stateful services cannot be supported due to asymmetric routing inherent in ECMP-based forwarding.

Even if you didn’t read documentation, but have networking skills, you’d know that protocols like NAT need to track network session state and even if you configure the same NAT rule on all of your ECMP-enabled edges, it won’t work, because due to ECMP, traffic can flow through one ESG on ingress and another ESG on egress. Since NAT tables are not synchronized, ESGs won’t be able to find the corresponding network flow in translation table and will drop the traffic.

ECMP and Firewall

But there’s another issue that doesn’t always come across or simply get forgotten about. You can deploy ESGs in ECMP mode, not configure any of the stateful services like VPN, NAT or LB, but still get network communication issues. Why? Because when you deploy an ESG, you always end up with firewall in enabled state. Firewall is also considered a stateful service.

From VVD 5.1 documentation:

SDDC-VISDN-032: For all ESGs deployed as ECMP North-South routers, disable the firewall. Use of ECMP on the ESGs is a requirement. Leaving the firewall enabled, even in allow all traffic mode, results in sporadic network connectivity. Services such as NAT and load balancing cannot be used when the firewall is disabled.

In fact, firewall is what actually tracks sessions and drops packets that don’t match existing network flows, not NAT itself. That’s also the reason why services like NAT and LB don’t work without firewall being enabled.

It often throws people off, because even having no rules in the firewall and setting default policy to accept will not prevent this issue from happening.

Demo

Here is a quick demonstration. I’m trying to establish an SSH session to a VM connected to a DLR behind two ESGs in ECMP mode.

I’m showing packet debug on both ESGs using the following command:

> debug packet display follow interface vNic_1 port_22

As you can see ingress traffic goes through E1 and egress traffic goes through E2:

E1: Packet Capture

E2: Packet Capture

Since session originated on E1, E2 interprets packets as invalid and immediately drops them:

From NSX Troubleshooting Guide:

Check for an incrementing value of a DROP invalid rule in the POST_ROUTING section of the show firewall command. Typical reasons include:

• Asymmetric routing issues
• …

Conclusion

It’s easy to end up in this situation, because firewall is enabled by default on a newly deployed ESG. And it’s hard to troubleshoot this issue, since it’s not quite obvious what’s actually going on unless you’ve already worked with ECMP before. So the best advice in this case is just to remember, if you want to use ECMP in NSX, make sure to disable firewall on ECMP-enabled ESGs. Use distributed firewall (DFW) instead.

Beginner’s Guide to HPE 5000 Series Switches

I don’t closely track the popularity of my blog. If what I share helps people in their day to day job, it’s already good enough to me. But I do look at site statistics now and then just out of curiosity and it seems that network-related posts get a lot of popularity. A blog post I wrote a while ago on Dell N4000 switches has quickly got in top five over the last year.

So it seems that there is a demand for entry-level switch configuration guides. I’ve worked with a quite a few different switch brands over the years, so I thought I will build on the success of the Dell blog post and this time write about HPE FlexNetwork/FlexFabric 5000 switch series.

Operating Systems

HPE has several network switch product lines. I won’t even try to cover all of them in this post. But it’s important to know that there are a few different operating systems you can encounter, while working with HPE network switches. There is a familiar ProCurve product portfolio (now merged with Aruba), which is based on ProVision operating system.

HPE FlexNetwork/FlexFabric 5000 series, on the other hand, is based on Comware operating system. It has a different CLI command set and can be a complete surprise if you’ve worked only with ProCurve switches before. So this blog post will be particularly valuable for those who’re dealing with HPE 5000 for the first time.

The following guide has been tested on a pair of HPE FlexFabric 5700-series switches. Even though commands are mostly the same, on other switch series, like FlexNetwork 5800, there might be some minor differences.

Initial Configuration

When the switch is booted for the first time it will start automatic configuration by trying to obtain settings over DHCP, which you can interrupt by Ctrl+C to get straight to CLI.

You start in user view where you can run display commands to review switch settings. To start the configuration, change to system view:

> system-view

Let’s start by configuring remote access to the switch. There are two ways you can do that. You either use the out-of-band management port:

> interface M-GigabitEthernet 0/0/0
> ip address 10.10.10.10 255.255.255.0
> ip route-static 0.0.0.0 0.0.0.0 10.10.10.1

Or you can configure a VLAN interface IP address:

> interface vlan-interface 1
> ip address 10.10.10.10 255.255.255.0
> ip route-static 0.0.0.0 0.0.0.0 10.10.10.1

Then configure switch name, enable SSH, set passwords and you can start managing the switch over SSH:

> sysname switchname

> public-key local create rsa
> ssh server enable
> user-interface vty 0 15
> authentication-mode scheme
> protocol inbound ssh

> super password simple yourpassword
> local-user admin
> password simple yourpassword
> authorization-attribute user-role level-0
> service-type ssh

User “admin” will have an unprivileged role. You will need to run the following command and enter password once logged in, to elevate to network admin rights:

> super

Intelligent Resilient Framework

In small non-business-critical environments one standalone switch is usually sufficient. In larger environments switches are typically deployed in pairs for redundancy. To simplify management and to avoid network loops most switches support some sort of MLAG or stacking. IRF is HPE’s version of it.

Determine what ports you’re going to use for IRF. There are two QSFP+ ports on 5700-series dedicated for it. And then on on the first switch (master) run the following commands (it’s recommended to shut down the ports before you set them up as IRF):

> irf member 1 priority 32
> int range FortyGigE 1/0/41 to FortyGigE 1/0/42
> shutdown
> irf-port 1/1
> port group interface FortyGigE 1/0/41
> irf-port 1/2
> port group interface FortyGigE 1/0/42
> int range FortyGigE 1/0/41 to FortyGigE 1/0/42
> undo shut
> save
> irf-port-configuration active

On the second switch (slave) run the following commands to change the IRF ID to 2:

> irf member 1 renumber 2
> reboot

When the switch comes up, configure IRF ports:

> irf member 2 priority 30
> int range FortyGigE 2/0/41 to FortyGigE 2/0/42
> shutdown
> irf-port 2/1
> port group interface FortyGigE 2/0/41
> irf-port 2/2
> port group interface FortyGigE 2/0/42
> int range FortyGigE 2/0/41 to FortyGigE 2/0/42
> undo shut
> save
> irf-port-configuration active

Now you can connect the physical IRF ports. IRF is a ring topology, that means (in my case) port 1/0/41 should connect to 2/0/42 and port 1/0/42 should connect to 2/0/41.

Second switch will automatically reboot and if all is configured correctly, you should see both switches join the IRF fabric. Member switch 1 has the highest priority of 32 and becomes the master:

> display irf

Firmware Upgrade

Firmware upgrade is the next logical step after you set up IRF. The latest firmware revision for the switches can be download from HPE web-site. Keep in mind you will need a HPE passport account, with a valid service agreement (SAID) added to it.

You will also need a TFTP server to upgrade the firmware. There are a few of them out there, but the most commonly used is probably Tftpd64.

When you get the TFTP server up and running and copy the firmware file to it, perform an upgrade:

> tftp 10.10.10.20 get 5700-CMW710-R2432P03.ipe
> boot-loader file flash:/5700-CMW710-R2432P03.ipe slot 1 main
> boot-loader file flash:/5700-CMW710-R2432P03.ipe slot 2 main
> irf auto-update enable
> reboot

Confirm firmware has been updated:

> display version

VLANs, Aggregation Groups and Tagging

In Comware the term “aggregation group” is used to describe what is a “port channel” in Cisco world. Trunk/access ports are also called tagged/untagged ports throughout the documentation.

In this section we will discuss a few common port configuration scenarios:

• Untagged ports, which can be your iSCSI storage array ports
• Tagged ports, such as your VMware host uplinks
• Aggregation groups, typically used for LAGs to upstream switches

First of all create all VLANs and give them descriptions:

> vlan 10
> description iSCSI
> vlan 20
> description Server
> vlan 30
> description Dev and test

Then specify untagged ports:

> vlan 10
> port te 1/0/1
> port te 2/0/1

To configure tagged ports and allow certain VLANs (ports will be added to the VLANs automatically):

> int te 1/0/2
> description ESX01 vmnic0
> port link-type trunk
> port trunk permit vlan 20 30
> int te 2/0/2
> description ESX02 vmnic0
> port link-type trunk
> port trunk permit vlan 20 30

And to create an LACP aggregation group:

> interface bridge-aggregation 1
> description Trunk to upstream switch
> link-aggregation mode dynamic
> port link-type trunk
> port trunk permit vlan 20 30

> interface te 1/0/3
> port link-aggregation group 1
> interface te 2/0/3
> port link-aggregation group 1

Common Commands

Other useful commands that don’t fall under any specific category, but handy to know.

Display switch configuration:

> display current-configuration

Save switch configuration:

> save

Shut down a port:

> int te 1/0/27
> shutdown

Undo a command:

> undo shutdown

Conclusion

Whether you are a network engineer new to the Comware operating system or a VMware administrator looking for a quick cheat sheet for FlexNetwork/FlexFabric switches, I hope this guide has helped you get the job done.

If this blog post gets the same amount of popularity, maybe it will turn into another series. But for now – over and out.

Dell Force10 Part 3: VLT Domain Configuration

In my previous post here I went through VLT basics and how it helps to establish a loop-free network topology in a modern datacenter. Now lets dive deeper and see how VLT is configured from FTOS CLI.

VLT Configuration

The first step is to configure the backup links and VLT interconnect. Dell S4048-ON switches have six 40Gb QSFP+ ports, two of which 1/49 and 1/50 we will use for VLTi. Repeat the same configuration on both switches.

# int range fo 1/49-1/50
# no shutdown

# interface port-channel 127
# description “VLT interconnect”
# channel-member fo 1/49
# channel-member fo 1/50
# no shutdown

Now that we have a VLT interconnect set up, let’s join the first switch to a VLT domain:

# vlt domain 1
# back-up destination 172.10.10.11
# peer-link port-channel 127
# primary-priority 1

First switch points to the second switch management IP for a backup destination, uses port channel 127 as a VLT interconnect and becomes a primary peer, because it’s given the lowest priority of 1.

Do the same on the second switch, but now point to the first switch management IP for backup and use the highest priority to make this switch a secondary peer:

# vlt domain 1
# back-up destination 172.10.10.10
# peer-link port-channel 127
# primary-priority 8192

To confirm the VLT state use the following command:

# sh vlt brief

As you can see, the VLTi and backup links are up and the switch can see its peer. For some additional VLT specific information use these commands:

# sh vlt statistics
# sh vlt backup-link

I would also recommend to use the following command to see the port channel state and confirm that both VLTi links are in up state:

# sh int po127

Conclusion

In this part of the Dell Force10 switch configuration series we quickly went through the initial VLT setup. We haven’t touched on VLT LAG configuration yet. We will take a closer look at it in the next blog post.

Dell Force10 Part 2: VLT Basics

Last time I made a blog post on initial configuration of Force10 switches, which you can find here. There I talked about firmware upgrade and basic features, such as STP and Flow Control. In this blog post I would like to touch on such a key feature of Force10 switches as Virtual Link Trunking (VLT).

VLT is Force10’s implementation of Multi-Chassis Link Aggregation Group (MLAG), which is similar to Virtual Port Channels (vPC) on Cisco Nexus switches. The goal of VLT is to let you establish one aggregated link to two physical network switches in a loop-free topology. As opposed to two standalone switches, where this is not possible.

You could say that switch stacking gives you similar capabilities and you would  be right. The issue with stacked switches, though, is that they act as a single switch not only from the data plane point of view, but also from the control plane point of view. The implication of this is that if you need to upgrade a switch stack, you have to reboot both switches at the same time, which brings down your network. If you have an iSCSI or NFS storage array connected to the stack, this may cause trouble, especially in enterprise environments.

With VLT you also have one data plane, but individual control planes. As a result, each switch can be managed and upgraded separately without full network downtime.

VLT Terminology

Virtual Link Trunking uses the following set of terms:

• VLT peer – one of the two switches participating in VLT (you can have a maximum of two switches in a VLT domain)
• VLT interconnect (VLTi) – interconnect link between the two switches to synchronize the MAC address tables and other VLT-related data
• VLT backup link – heartbeat link to send keep alive messages between the two switches, it’s also used to identify switch state if VLTi link fails
• VLT – this is the name of the feature – Virtual Link Trunking, as well as a VLT link aggregation group – Virtual Link Trunk. We will call aggregated link a VLT LAG to avoid ambiguity.
• VLT domain – grouping of all of the above

VLT Topology

This’s what a sample VLT domain looks like. S4048-ON switches have six 40Gb QSFP+ ports, two of which we use for a VLT interconnect. It’s recommended to use a static LAG for VLTi.

Two 1Gb links are used for VLT backup. You can use switch out-of-band management ports for this. Four 10Gb links form a VLT LAG to the upstream core switch.

Use Cases

So where is this actually helpful? Vast majority of today’s environments are virtualized and do not require LAGs. vSphere already uses teaming on vSwitch uplinks for traffic distribution across all network ports by default. There are some use cases in VMware environments, where you can create a LAG to a vSphere Distributed Switch for faster link failure convergence or improved packet switching. Unless you have a really large vSphere environment this is generally not required, but you may use this option later on if required. Read Chris Wahl’s blog post here for more info.

Where VLT is really helpful is in building a loop-free network topology in your datacenter. See, all your vSphere hosts are connected to both Force10 switches for redundancy. Since traffic comes to either of the switches depending on which uplink is being picked on a ESXi host, you have to make sure that VMs on switch 1 are able to communicate to VMs on switch 2. If all you had in your environment were two Force10 switches, you would establish a LAG between the two and be done with it. But if your network topology is a bit larger than this and you have at least a single additional core switch/router in your environment you’d be faced with the following dilemma. How can you ensure efficient traffic switching in your network without creating loops?

You can no longer create a LAG between the two Force10 switches, as it will create a loop. Your only option is to keep switches connected only to the core and not to each other. And by doing that you will cause all traffic from VMs on switch 1 destined to VMs on switch 2 and vise versa to traverse the core.

And that’s where VLT comes into play. All east-west traffic between servers is contained within the VLT domain and doesn’t need to traverse the core. As shown above, if we didn’t use VLT, traffic from one switch to another would have to go from switch 1 to core and then back from core to switch 2. In a VLT domain traffic between the switches goes directly form switch 1 to switch 2 using VLTi.

Conclusion

That’s a brief introduction to VLT theory. In the next few posts we will look at how exactly VLT is configured and map theory to practice.

Dell Force10 Part 1: Initial Configuration

When it comes to networking Dell has two main series of switches. PowerConnect/N-series, which run DNOS 6.x operating system. And S/Z-series switches, which run on DNOS 9.x derived from Force10 OS (FTOS). In this series of blogs we will go through the configuration of Force10 switch series and use Dell S4048-ON top of the rack switch as an example.

Interesting to note, that unlike other S-series switches S4048-ON is an Open Networking switch. Dell is one of the first companies which apart from its own OS lets customers run other operating systems on its network switches, such as Cumulus Linux OS and Big Switch Networks Switch Light OS. While Cumulus and Big Switch has its own use cases, in this blog we will look specifically at configuring FTOS.

Boot process

S4048-ON comes from the factory pre-configured for bare metal provisioning (BMP). This is what you will see when you boot the switch for the first time:

If you just want to boot FTOS, simply skip the BMP by choosing A and switch will boot the OS.

After some time BMP will time out. If you’ve missed the above wizard, you can also disable BMP from CLI using the following commands:

> enable
# stop bmp
# config
# reload-type normal-reload
# exit
# reload

When prompted choose to save the configuration and proceed with reload. After the switch has rebooted check that the next boot is set to normal reload:

# show reload-type

Initial configuration

First steps of any switch installation is assigning a hostname and management interface settings:

# hostname DELL4048-SWITCH
# int managementethernet 1/1
# ip address 172.10.10.2/24
# no shut
# management route 0.0.0.0/0 172.10.10.10

Then set admin / enable passwords and allow remote management via SSH:

# enable password 123456
# username admin password 123456
# ip ssh server enable

Configure time zone and NTP:

# clock timezone UTC 11
# ntp server 172.10.10.20
# show ntp associations
# show ntp status
# show clock

Firmware upgrade

Force10 switches have two boot banks A: and B:. It’s a good practice to upload new firmware into one boot bank and keep the old firmware in the other in case you need to roll back.

The easiest way to upgrade is via TFTP using Tftpd64, which you can download for free from here. If you’re upgrading an existing switch, make sure to save the running config and make a backup. If it’s an initial install you can skip this step.

# copy run start
# copy start tftp://10.0.0.1/FORCE10_SWITCH_01.01.16.conf

Then upload new firmware to image B:, change active boot bank to B: and reload:

# show version
# show boot system stack-unit 1
# upgrade system tftp://10.0.0.1/FTOS-SK-9.9.0.0P9.bin b:
# conf t
# boot system stack-unit 1 primary system b:
# exit
# reload

You will be prompted to save the configuration and reboot. After the reboot you may be asked to enable SupportAssist. SuppotAssist helps to automatically open Dell service tickets if there is a switch fault. You can enable SupportAssist by running the following commands and answering prompts:

# conf t
# support-assist activate
# support-assist activity full-transfer start now
# show support-assist status

My pair of switches were configured in a Virtual Link Trunking (VLT) domain. I’ll explain how VLT works later in the series. But from the upgrade point of view, each switch in a VLT domain is treated as a separate switch and has to be upgraded separately. If you decided to use a stack instead of VLT, you can find the upgrade process for a Force10 stack in my other post about Dell MXL switches here.

Spanning tree

Spanning Tree Protocol (STP) helps to prevent network topology loops and is highly recommended for use in any network. Switches connected in an actual loop topology in today’s networks are rare. But STP can save you from consequences of a potential human error, such as port channel misconfiguration. If instead of creating one port channel with two links, you by mistake create two port channels with one link each and both carry the same VLANs, you’ve accidentally created a loop, which will bring your whole network to an immediate halt.

It’s a good practice to enable STP as a safeguard mechanism from such configuration errors. S4048-ON supports STP, RSTP, MSTP and PVST+. In my case S4048s were uplinked into HP core, which supported STP, RSTP and MSTP. If you have Cisco switches in your network core you can use PVST+. In my case I used RSTP, which is a good choice if you don’t require enhancements of MSTP and PVST+ in your network. Just make sure to not use the basic STP protocol, as it provides the slowest convergence.

# protocol spanning-tree rstp
# no disable
# show spanning-tree rstp

In every STP topology there is also a root switch, which by default is selected automatically. For a more deterministic STP behaviour it’s recommended to select the root switch manually, by assigning the lowest STP priority to it. Typically your core switch should be your root switch. In my case it was a HP core switch, which was assigned priority of “0”.

When configuring server and storage facing ports make sure to enable EdgePort mode to minimize the time it takes for the port to come online:

# int range Te1/45-1/48
# spanning-tree rstp edge-port
# switchport
# no shut

If you want to know more about how STP works, you can read a few of my previous blog posts on STP here and here.

Flow control

To avoid dropped packets on 10Gb switch ports at times of potential heavy utilization it is also a best practice to as a minimum enable bi-directional Flow Control on the storage array ports. I enabled it on the iSCSI links connected from the Dell Compellent storage array:

# int range Te1/17-1/18
# flowcontrol rx on tx on

If you specifically interested in switch best practices for Compellent and EqualLogic storage arrays, Dell has a full list of guides for various switches at communitites wiki here.

Port channels and VLANs

Port channels and VLANs are configured similarly to any other switch, but I include them here in case you want to know the syntax. In this example we have two access ports 1/46 and 1/47 and an uplink to the core configured as port channel 1:

# interface port-channel 1
# switchport
# no shutdown

# interface range Te1/1-1/2
# port-channel-protocol LACP
# port-channel 1 mode active
# no shutdown

# int vlan 254
# untagged Te1/46-1/47
# tagged po 1

Keep in mind, that port channels are used either in one switch configurations or when two or more switches are stacked together. If you’re using Virtual Link Trunking (VLT), you will need to create Virtual Link Trunks (VLTs). Which are similar to port channels, but have a slightly different syntax. We will talk about VLT in much more detail in the following Force10 blogs.

Conclusion

One feature which I didn’t specifically mentioned in this blog post was Jumbo Frames. I tend not to use it in my deployments until I see convincing evidence of it making a difference for iSCSI/NFS storage implementations. I did a post about Jumbo Frames long time ago here and hasn’t changed my opinion ever since. Interested to here your thoughts if have a different take on that.

References

• My blog:
  • Jumbo Frames justified?
  • Spanning Tree Protocol Overview
  • How STP and RSTP converge
  • Force10 MXL: Firmware Upgrade

• External sources:
  • Switch Configuration Guides for PS Series or SC Series SANs
  • Dell Networking S4048-ON: Switch Configuration Guide for Dell SC Series iSCSI SANs

Force10 and vSphere vDS Interoperability Issue

Recently I had an opportunity to work with Dell FX2 platform from the design and delivery point of view. I was deploying a FX2s chassis with FC630 blades and FN410S 10Gb I/O aggregators.

I ran into an interesting interoperability glitch between Force10 and vSphere distributed switch when using LLDP. LLDP is an equivalent of Cisco CDP, but is an open standard. And it allows vSphere administrators to determine which physical switch port a given vSphere distributed switch uplink is connected to. If you enable both Listen and Advertise modes, network administrators can get similar visibility, but from the physical switch side.

In my scenario, when LLDP was enabled on a vSphere distributed switch, uplinks on all ESXi hosts started disconnecting and connecting back intermittently, with log errors similar to this:

Lost uplink redundancy on DVPorts: “1549/03 4b 0b 50 22 3f d7 8f-28 3c ff dd a4 76 26 15”, “1549/03 4b 0b 50 22 3f d7 8f-28 3c ff dd a4 76 26 15”, “1549/03 4b 0b 50 22 3f d7 8f-28 3c ff dd a4 76 26 15”, “1549/03 4b 0b 50 22 3f d7 8f-28 3c ff dd a4 76 26 15”. Physical NIC vmnic1 is down.

Network connectivity restored on DVPorts: “1549/03 4b 0b 50 22 3f d7 8f-28 3c ff dd a4 76 26 15”, “1549/03 4b 0b 50 22 3f d7 8f-28 3c ff dd a4 76 26 15”. Physical NIC vmnic1 is up

Uplink redundancy restored on DVPorts: “1549/03 4b 0b 50 22 3f d7 8f-28 3c ff dd a4 76 26 15”, “1549/03 4b 0b 50 22 3f d7 8f-28 3c ff dd a4 76 26 15”, “1549/03 4b 0b 50 22 3f d7 8f-28 3c ff dd a4 76 26 15”, “1549/03 4b 0b 50 22 3f d7 8f-28 3c ff dd a4 76 26 15”. Physical NIC vmnic1 is up

Issue Troubleshooting

FX2 I/O aggregator logs were reviewed for potential errors and the following log entries were found:

%STKUNIT0-M:CP %DIFFSERV-5-DSM_DCBX_PFC_PARAMETERS_MISMATCH: PFC Parameters MISMATCH on interface: Te 0/2

%STKUNIT0-M:CP %IFMGR-5-OSTATE_DN: Changed interface state to down: Te 0/2

%STKUNIT0-M:CP %IFMGR-5-OSTATE_UP: Changed interface state to up: Te 0/2

This clearly looks like some DCB negotiation issue between Force10 and the vSphere distributed switch.

Root Cause

Priority Flow Control (PFC) is one of the protocols from the Data Center Bridging (DCB) family. DCB was purposely built for converged network environments where you use 10Gb links for both Ethernet and FC traffic in the form of FCoE. In such scenario, PFC can pause Ethernet frames when FC is not having enough bandwidth and that way prioritise the latency sensitive storage traffic.

In my case NIC ports on Qlogic 57840 adaptors were used for 10Gb Ethernet and iSCSI and not FCoE (which is very uncommon unless you’re using Cisco UCS blade chassis). So the question is, why Force10 switches were trying to negotiate FCoE? And what did it have to do with enabling LLDP on the vDS?

The answer is simple. LLDP not only advertises the port numbers, but also the port capabilities. Data Center Bridging Exchange Protocol (DCBX) uses LLDP when conveying capabilities and configuration of FCoE features between neighbours. This is why enabling LLDP on the vDS triggered this. When Force10 switches determined that vDS uplinks were CNA adaptors (which was in fact true, I was just not using FCoE) it started to negotiate FCoE using DCBX. Which didn’t really go well.

Solution

The easiest solution to this problem is to disable DCB on the Force10 switches using the following command:

# conf t
# no dcb enable

Alternatively you can try and disable FCoE from the ESXi end by using the following commands from the host CLI:

# esxcli fcoe nic list
# esxcli fcoe nic disable -n vmnic0

Once FCoE has been disabled on all NICs, run the following command and you should get an empty list:

# esxcli fcoe adapter list

Conclusion

It is still not clear why PFC mismatch would cause vDS uplinks to start flapping. If switch cannot establish a FCoE connection it should just ignore it. Doesn’t seem to be the case on Force10. So if you run into a similar issue, simply disable DCB on the switches and it should fix it.

Beginner’s Guide to Dell N4000 Series Switches

Dell N-Series switches run on Dell Network Operating System (DNOS) version 6.x. Unlike Dell S-Series switches which run on DNOS 9.x, derived from  Force10 Operation System (FTOS), DNOS 6.x came from the PowerConnect switch series and share the same codebase. So if you’ve ever worked with PowerConnect switches, N-Series syntax should be very familiar.

In my case I had two Dell N4032F switches. But the same set of commands applies to any other N4000 Series switch.

Initial Configuration

When you first turn the switch on, it gives you 60 seconds to enter the wizard, where you can set up network settings for the Out-of-Band (OOB) management interface and change the admin password. If you miss it you can reboot the switch and it will show the same wizard prompt again when it boots up. Or you can set it up from the CLI:

# interface out-of-band
# ip address 10.10.10.10 255.255.255.0 10.10.10.254

# show ip interface out-of-band

Once you get to the CLI prompt, configure hostname and enable SSH:

# hostname n4032f-prod

# crypto key generate rsa
# crypto key generate dsa
# ip ssh server
# ip telnet server disable

Stacking

Dell N4000 Series switches support both stacking and MLAG (Multi-chassis Link Aggregation). One of the drawbacks of the stack configuration is disruptive firmware upgrades. When you update firmware on the stack master, firmware is distributed to all stack members and all switches are rebooted simultaneously.

In MLAG each switch has its own Control Plane and can be rebooted independently. Which is MLAG’s shortcoming at the same time, because unlike stack, where all units act as one switch, in MLAG you have to manage each switch separately.

In my case I chose stacking for its simplicity.

N4000 switches are stacked using the two 40Gb QSFP ports located at the front. QSFP ports are not configured in stack mode by default. Which you need to change on both switches before you can build a stack:

# stack
# stack-port Fortygigabitethernet 1/1/1 stack
# stack-port Fortygigabitethernet 1/1/2 stack

# show switch stack-ports

Once QSFP ports on both switches are configured, disconnect power from both switches and boot the switch you want to be the stack master first (typically the top switch). When the first switch has fully booted, boot the second switch and check the status. This is what you should see:

# show switch

Firmware Upgrade

If it’s not a brand new switch, save the config before doing the firmware upgrade:

# copy run start
# copy running-config tftp://10.10.10.100/backup.txt

You can use any TFTP server for the firmware upgrade, such as the free Tftpd64 server.

Then you upload the firmware image to the stack master and reload the stack:

# copy tftp://10.10.10.100/N4000v6.2.7.2.stk backup
# boot system backup
# reload
# show version

Firmware is uploaded to a backup image. Then you select the backup image for the next boot and reload the stack. When both switches reboot you should see something similar to this:

As part of the upgrade process the new firmware is automatically uploaded from the master to all stack members, which is a default behaviour. You can confirm it is enabled using the following command:

# show auto-copy-sw

Flow Control, Jumbo Frames and iSCSI Optimization

In my case I used two N4032F switches for an iSCSI backbone, so I needed to make sure that Flow Control and Jumbo Frames are enabled on the switch.

Flow Control is enabled by default, which you can confirm by the following command:

# show storm-control

To globally enable Jumbo Frames on all ports type:

# system jumbo mtu 9216

# show system mtu

Interestingly, Dell N4000 Series switches also have built-in iSCSI optimization, which can detect iSCSI sessions by snooping the traffic on ports 3260 and 860. It then prioritizes iSCSI traffic over the other types of traffic to guarantee low latency for storage I/O. To show iSCSI settings:

# show iscsi

By default switches only track the sessions. Traffic prioritization is disabled by default and has to be enabled manually. This didn’t matter in my case, as the switches were dedicated for storage traffic. But if you share switches between storage and server traffic, you may want to enable it. Refer to the switch User’s Configuration Guide for details.

If you’re using a Dell Compellent storage array with N4000 switches, also make sure to apply a Compellent profile to the ports where storage array is connected to:

# macro global apply profile-compellent-nas $interface_name te1/0/1
# macro global apply profile-compellent-nas $interface_name te1/0/2
# macro global apply profile-compellent-nas $interface_name te1/0/3
# macro global apply profile-compellent-nas $interface_name te1/0/4

VLANs, Trunks and Port Channels

Again, I didn’t use any VLANs and Trunks, because switches were dedicated for iSCSI traffic and were separate from the LAN core. And I didn’t need Port Channels either, as they are not required for iSCSI.

Your scenario might be different. For instance, if you have vSphere hosts connected to a NetApp array over NFS, you may want to create a Multi-Mode (LACP) VIF on the NetApp side. If that’s the case, to create a port channel on the Multi-Mode VIF ports use the following:

# interface range te1/0/2,te2/0/2
# channel-group 1 mode active
# show intefaces po1

If the switches are used for both storage and VM traffic, then you’ll need to configure the server ports and uplink them to your network core. Create your VLANs first:

# vlan 10,20,30

Configure vSwitch uplinks from the ESXi hosts. In a typical vSphere environment, traffic is tagged on the vSwitch side, which means that server ports should be configured as trunks:

# interface range te1/0/3-6,te2/0/3-6
# switchport mode trunk
# switchport trunk allowed vlan 10,20,30

And finally configure uplinks to the network core. Depending on how your LAN core is set up, you may want to create a port channel to the upstream switch and trunk the required VLANs:

# interface range te1/0/1,te2/0/1
# channel-group 2 mode active
# switchport mode trunk
# switchport trunk allowed vlan 10,20,30
# show intefaces po2

Conclusion

This guide didn’t include information on Spanning Tree, QoS or any of the switch Layer 3 features, but I hope it could get you started. At the end of the day, every environment is different. If you need additional information refer to the following guides from the Dell web-site:

• Stacking Dell Networking Switches: N4032, N4032F, N4064, N4064F
• Using MLAG in Dell Networks
• Dell Networking N4000 Series Switch Getting Started Guide
• Dell Networking N-Series N1500, N2000, N3000, and N4000 Switches User’s Configuration Guide
• Dell Networking N-Series N1500, N2000, N3000, and N4000 Switches CLI Reference Guide

 

Traffic Load Balancing in Cisco UCS

Whenever I deploy a Cisco UCS at a customer the question I get asked a lot is how traffic flows within the system between VMs running on the blades and FEX modules, FEX modules and Fabric Interconnects and finally how it’s uplinked to the network core.

Cisco has a range of CNA cards for UCS blades. With VIC 1280 you get 8 x 10Gb ports split between two FEX modules for redundancy. And FEX modules on their own can have up to 8 x 10Gb Fabric Interconnect facing interfaces, which can give you up to 160Gb of bandwidth per chassis. And all these numbers may sound impressive, but unless you understand how your VMs traffic flows through UCS it’s easy to make wrong assumptions on what per VM and aggregate bandwidth you can achieve. So let’s dive deep into UCS and shed some light on how VM traffic is load-balanced within the system.

UCS Hardware Components

Each Fabric Extender (FEX) has external and internal ports. External FEX ports are patched to FIs and internal ports are internally wired to the blade adapters. FEX 2204 has 4 external and 16 internal and FEX 2208 has 8 external and 32 internal ports.

External ports are connected to FIs in powers of two: 1, 2, 4 or 8 ports per FEX and form a port channel (make sure to use “Port Channel” link grouping preference under Chassis/FEX Discovery Policy). Same rule is applied to blade Virtual Interface Cards (VIC). The most common VIC 1240 and 1280 have 4 x 10Gb and 8 x 10Gb ports respectively and also form a port channel to the internal FEX ports. Every VIC adaptor is connected to both FEX modules for redundancy.

Fabric Interconnects are then patched to your network core and FC Fabric (if you have one). Whether Ethernet uplinks will be individual uplinks or port channels will depend on your network topology. For fibre uplinks the rule of thumb is to patch FI A to your FC Fabric A and FI B to FC Fabric B, which follows the common FC traffic isolation principle.

Virtual Circuits

To provide network and storage connectivity to blades you create virtual NICs and virtual HBAs on each blade. Since internally UCS uses FCoE to transfer FC frames, both vNICs and vHBAs use the same 10GbE uplinks to send and receive traffic. Worth mentioning that Cisco uses Data Center Bridging (DCB) protocol with it’s sub-protocols Priority Flow Control (PFC) and Enhanced Transmission Selection (ETS), which guarantee that FC frames have higher priority in the queue and are processed first to ensure low latency. But I digress.

UCS assigns a virtual circuit to each virtual adaptor, which is a representation of how the traffic traverses the system all the way from the VIC port to a FEX internal port, then FEX external port, FI server port and finally a FI uplink. You can trace the full path of each virtual adaptor in UCS Manager by selecting a Service Profile and viewing the VIF Paths tab.

In this example we have a blade with four vNICs and two vHBAs which are split between two fabrics. All virtual adaptors on fabric A are connected through VIC port channel PC-1283 which is represented as port channel PC-1025 on the FEX A side. Then traffic leaves FEX A and reaches the Fabric Interconnect A which sends the traffic out to the network core through port channel A/PC-1.

You can also get the list of port channels from the FI CLI:

# connect nxos
# show port-channel summary

Network Load Balancing

Now that we know how all components are interconnected to each other, let’s discuss the traffic flow in a typical VMware environment and how we achieve the massive network throughput that UCS provides.

As an example let’s take a look at the vSwitch where your VM Network port group is configured. vSwitch will have two uplinks – one goes to Fabric A and the other one to Fabric B for redundancy. Default load balancing policy on a vSwitch is “Route based on the originating port ID”, which essentially pins all traffic for a VM to a particular uplink. vSphere makes sure that VMs are evenly distributed between the uplinks to use all network bandwidth available.

From each uplink (or vNIC in UCS world) traffic is forwarded through an adapter port channel to a FEX, then to a Fabric Interconnect and leaves UCS from a FI uplink. Within UCS traffic is distributed between port channel members using source/destination IP hash algorithm. Which is even more granular and is capable of very efficient traffic distribution between all members of a port channel all the way up to your network core.

If you look at the vSwitch you’ll see that with UCS each uplink shows the maximum available bandwidth from vNIC and is not limited to a port channel member speed of 10Gb. Why is this so powerful? Because with UCS you don’t need to slice adapter’s available bandwidth between different types of traffic. Even though you provision multiple vNICs and vHBAs for the vSphere hosts, UCS uses the same port channel links (20Gb in the example below) from the VIC adapter to transfer all traffic and takes care of load balancing for you.

You may legitimately ask, if UCS uses the same pipe to transfer all data regardless of which vSwitch uplink is being used, then how can I make sure that different types of traffic, such as vMotion, storage, VM traffic, replication, etc, do not compete for the same pipe? First you need to ask yourself if you can saturate that much bandwidth with your workloads. If the answer is yes, then you can use another great feature available in UCS, which is QoS. QoS lets you assign a minimum available bandwidth guarantee on a per vNIC/vHBA basis. But that’s a topic for another blog post.

References

In this post I tried to summarise the logic behind UCS traffic distribution. If you want to dig deeper in UCS network architecture, then there’re a lot of great bloggers out there. I would like to call out the following authors:

• Brad Hedlund – if you want to better understand overall UCS architecture and network best practices, Bred has a great series of videos:
  • Cisco UCS Networking videos (in HD), Updated & Improved!
• Colin Lynch (UCSguru) – this blog is dedicated specifically to UCS. If you want to know more about the virtual circuits, I would recommend these two blog posts:
  • Understanding UCS VIF Paths
  • Under the Cisco UCS Kimono
• Matthew Oswalt – Keeping It Classless is a blog which is geared more towards DevOps, but also has a few UCS articles, I found this one particularly useful:
  • Cisco UCS Port-Channeling

 

Force10 MXL: Firmware Upgrade

Uploading new firmware image

MXL switches keep two firmware images – A and B. You can set either one of them to be active. Use the following command to list firmware version of all stack members and see which one is active:

 # show boot system stack-unit all

To upload firmware you will need a TFTP server. You can use TFTPD64 (also called TFTPD32), which can be downloaded from Philippe Jounin page here .

If the active firmware image is in A, upload new firmware to B. You’ll also be asked if you want to upload firmware to all switches in the stack.

# upgrade system tftp://10.0.0.1/FTOS-XL-9.5.0.0P2.bin b:

At the time of upgrade the latest version was 9.7. This version was 2 weeks old and wasn’t recommended for use in production. Version 9.6 had a major bug. As a result version 9.5SP2 was chosen for the upgrade.

Double-check that new firmware has been successfully distributed to all switches:

 # show boot system stack-unit all

Backing up startup config

Make sure you do not miss this step. If something goes wrong and switch looses its config, you’ll have to recreate all configuration from scratch. Imagine the consequences.

# copy start tftp://10.0.0.1/MXL_01.01.15.conf

Reloading the stack

Once firmware is uploaded and config is saved, reload the stack. Be mindful that it’s a disruptive procedure and all links connected to the stack will go down. A reboot shouldn’t take more than a couple of minutes, but make sure you do that after hours.

# conf t
# boot system stack-unit all primary system b:
# exit
# copy run start
# reload

Confirm that the active firmware image is now B. And that concludes the switch stack upgrade.