Installation Instructions for OpenFlow Reference Release This document describes how to build, install, and execute the reference implementation of OpenFlow. Please send any comments to: Prerequisites ------------- To compile the userspace programs in the OpenFlow reference distribution, you will need the following software: - A make program, e.g. GNU make (http://www.gnu.org/software/make/). BSD make should also work. - The GNU C compiler (http://gcc.gnu.org/). We generally test with version 4.1 or 4.2. - libssl, from OpenSSL (http://www.openssl.org/), is optional but recommended. libssl is required to establish confidentiality and authenticity in the connections among OpenFlow switches and controllers. To enable, compile with --enable-ssl=yes If you are working from a Git tree or snapshot (instead of from a distribution tarball), or if you modify the OpenFlow build system, you will also need the following software: - Autoconf version 2.59 or later (http://www.gnu.org/software/autoconf). - Automake version 1.10 or later (http://www.gnu.org/software/automake). - pkg-config (http://pkg-config.freedesktop.org/wiki/). We test with version 0.22. The optional Linux module has additional prerequisites, described later in the section "Building and Testing the Linux Kernel-Based Switch". Building Userspace Programs --------------------------- The OpenFlow distribution includes two implementations of the switch: one entirely in userspace, for portability and ease of installation, and another with a Linux kernel module component that is more difficult to install but should also yield better performance. These instructions describe how to build the userspace components of the OpenFlow distribution. Refer to "Building and Testing the Linux Kernel-Based Switch", below, for additional instructions on how to build the optional Linux kernel module. 1. In the top source directory, configure the package by running the configure script. You can usually invoke configure without any arguments: % ./configure To use a specific C compiler for compiling OpenFlow user programs, also specify it on the configure command line, like so: % ./configure CC=gcc-4.2 The configure script accepts a number of other options and honors additional environment variables. For a full list, invoke configure with the --help option. 2. Run make in the top source directory: % make The following binaries will be built: - Switch executable: switch/switch. This executable is built only if the configure script detects a supported interface to network devices. Refer to README for a list of OSes whose network device interfaces are supported. - Secure channel executable: secchan/secchan. - Controller executable: controller/controller. - Datapath administration utility: utilities/dpctl. - Runtime logging configuration utility: utilities/vlogconf. 3. (Optional) Run "make install" to install the executables and manpages into the running system, by default under /usr/local. Testing Userspace Programs -------------------------- 0. The commands below must run as root, so log in as root, or use a program such as "su" to become root temporarily. 1. Start the OpenFlow controller running in the background, by running the "controller" program with a command like the following: # controller ptcp: & This command causes the controller to bind to port 975 (the default) awaiting connections from OpenFlow switches. See controller(8) for details. 2. On the same machine, use the "switch" program to start an OpenFlow switch, specifying network devices to use as switch ports on the -i option as a comma-separated list, like so: # switch tcp:127.0.0.1 -i eth1,eth2 The network devices that you specify should not have configured IP addresses. 3. The controller causes each switch that connects to it to act like a learning Ethernet switch. Thus, devices plugged into the specified network ports should now be able to send packets to each other, as if they were plugged into ports on a conventional Ethernet switch. Troubleshooting: if the commands above do not work, try using the -v or --verbose option on the controller or switch commands, which will cause a large amount of debug output from each program. Remote switches: These instructions assume that the controller and the switch are running on the same machine. This is an easy configuration for testing, but a more conventional setup would run a controller on one machine and one or more switches on different machines. To do so, simply specify the IP address of the controller as the first argument to the switch program (in place of 127.0.0.1). (Note: The userspace switch must be connected to the controller over a "control network" that is physically separate from the one that the switch and controller are controlling. The kernel-based switch does not have this limitation.) Secure operation over SSL ------------------------- The instructions above set up OpenFlow for operation over a plaintext TCP connection. Production use of OpenFlow should use SSL[*] to ensure confidentiality and authenticity of traffic among switches and controllers. The source must be configured with --enable-ssl=yes to build with SSL support. To use SSL with OpenFlow, you must set up a public-key infrastructure (PKI) including a pair of certificate authorities (CAs), one for controllers and one for switches. If you have an established PKI, OpenFlow can use it directly. Otherwise, refer to "Establishing a Public Key Infrastructure" below. To configure the controller to listen for SSL connections on port 976 (the default), invoke it as follows: # controller -v pssl: --private-key=PRIVKEY --certificate=CERT \ --ca-cert=CACERT where PRIVKEY is a file containing the controller's private key, CERT is a file containing the controller CA's certificate for the controller's public key, and CACERT is a file containing the root certificate for the switch CA. If, for example, your PKI was created with the instructions below, then the invocation would look like: # controller -v pssl: --private-key=ctl-privkey.pem \ --certificate=ctl-cert.pem --ca-cert=pki/switchca/cacert.pem To configure a switch to connect to a controller running on port 976 (the default) on host 192.168.1.2 over SSL, invoke it as follows: # switch -v ssl:192.168.1.2 -i INTERFACES --private-key=PRIVKEY \ --certificate=CERT --ca-cert=CACERT where INTERFACES is the command-separated list of network device interfaces, PRIVKEY is a file containing the switch's private key, CERT is a file containing the switch CA's certificate for the switch's public key, and CACERT is a file containing the root certificate for the controller CA. If, for example, your PKI was created with the instructions below, then the invocation would look like: # secchan -v -i INTERFACES ssl:192.168.1.2 --private-key=sc-privkey.pem \ --certificate=sc-cert.pem --ca-cert=pki/controllerca/cacert.pem [*] To be specific, OpenFlow uses TLS version 1.0 or later (TLSv1), as specified by RFC 2246, which is very similar to SSL version 3.0. TLSv1 was released in January 1999, so all current software and hardware should implement it. Establishing a Public Key Infrastructure ---------------------------------------- If you do not have a PKI, the ofp-pki script included with OpenFlow can help. To create an initial PKI structure, invoke it as: % ofp-pki new-pki which will create and populate a new directory named "pki" under the current directory. The pki directory contains two important subdirectories. The controllerca subdirectory contains controller certificate authority related files, including the following: - cacert.pem: Root certificate for the controller certificate authority. This file must be provided to the switch or secchan program with the --ca-cert option to enable it to authenticate valid controllers. - private/cakey.pem: Private signing key for the controller certificate authority. This file must be kept secret. There is no need for switches or controllers to have a copy of it. The switchca subdirectory contains switch certificate authority related files, analogous to those in the controllerca subdirectory: - cacert.pem: Root certificate for the switch certificate authority. This file must be provided to the controller program with the --ca-cert option to enable it to authenticate valid switches. - private/cakey.pem: Private signing key for the switch certificate authority. This file must be kept secret. There is no need for switches or controllers to have a copy of it. After you create the initial structure, you can create keys and certificates for switches and controllers with ofp-pki. To create a controller private key and certificate in files named ctl-privkey.pem and ctl-cert.pem, for example, you could run: % ofp-pki req+sign ctl controller ctl-privkey.pem and ctl-cert.pem would need to be copied to the controller for its use at runtime (they could then be deleted from their original locations). The --private-key and --certificate options of controller, respectively, would point to these files. Analogously, to create a switch private key and certificate in files named sc-privkey.pem and sc-cert.pem, for example, you could run: % ofp-pki req+sign sc switch sc-privkey.pem and sc-cert.pem would need to be copied to the switch for its use at runtime (they could then be deleted from their original locations). The --private-key and --certificate options, respectively, of switch and secchan would point to these files. Building and Testing the Linux Kernel-Based Switch -------------------------------------------------- The OpenFlow distribution also includes a Linux kernel module that can be used to achieve higher switching performance at a cost in portability and ease of installation. Compiling the kernel module has the following prerequisites in addition to those listed in the "Prerequisites" section above: - A supported Linux kernel version. Please refer to README for a list of supported versions. The OpenFlow datapath requires bridging support (CONFIG_BRIDGE) to be built as a kernel module. (This is common in kernels provided by Linux distributions.) The bridge module must not be loaded or in use. If the bridge module is running (check with "lsmod | grep bridge"), you must remove it ("rmmod bridge") before starting the datapath. - The correct version of GCC for the kernel that you are building the module against: * To build a kernel module for a Linux 2.6 kernel, you need the same version of GCC that was used to build that kernel (usually version 4.0 or later). * To build a kernel module for a Linux 2.4 kernel, you need an earlier version of GCC, typically GCC 2.95, 3.3, or 3.4. - A kernel build directory corresponding to the Linux kernel image the module is to run on. Under Debian and Ubuntu, for example, each linux-image package containing a kernel binary has a corresponding linux-headers package with the required build infrastructure. To build the kernel module, follow the build process described under "Building Userspace Programs" above, but pass the location of the kernel build directory as an additional argument to the configure script, as described under step 1 in that section. Specify the location on --with-l26 for Linux 2.6, --with-l24 for Linux 2.4. For example, to build for a running instance of Linux 2.6: % ./configure --with-l26=/lib/modules/`uname -r`/build To build for a running instance of Linux 2.4: % ./configure --with-l24=/lib/modules/`uname -r`/build If you wish to build OpenFlow for an architecture other than the architecture used for compilation, you may specify the kernel architecture string using the KARCH variable when invoking the configure script. For example, to build OpenFlow for MIPS with Linux 2.4: % ./configure --with-l24=/path/to/linux-2.4 KARCH=mips If you have hardware that supports accelerated OpenFlow switching, and you have obtained a hardware table module for your hardware and extracted it into the OpenFlow reference distribution source tree, then you may also enable building support for the hardware switching table with --enable-hw-tables. For example, if your hardware switching table is in a directory named datapath/hwtable-foomatic, you could compile support for it with the running Linux 2.6 kernel like so: % ./configure --with-l26=/lib/modules/`uname -r`/build \ --enable-hw-tables=foomatic For more information about hardware table modules, please read README.hwtables at the root of the OpenFlow distribution tree. In addition to the binaries listed under step 2 in "Building Userspace Programs" above, "make" will build the following kernel modules: datapath/linux-2.6/openflow_mod.ko (if --with-l26 was specified) datapath/linux-2.4/openflow_mod.o (if --with-l24 was specified) "make" will also build a kernel module for each hardware switch table enabled with --enable-hw-tables. Once you have built the kernel modules, activating them requires only running "insmod", e.g.: (Linux 2.6) % insmod datapath/linux-2.6/openflow_mod.ko (Linux 2.4) % insmod datapath/linux-2.4/compat24_mod.o % insmod datapath/linux-2.4/openflow_mod.o After you load the openflow module, you may load one hardware switch table module (if any were built) to enable support for that hardware switching table. The insmod program must be run as root. You may need to specify a full path to insmod, which is usually in the /sbin directory. To verify that the modules have been loaded, run "lsmod" (also in /sbin) and check that openflow_mod appears in the result. Testing the Kernel-Based Implementation --------------------------------------- The OpenFlow kernel module must be loaded, as described in the previous section, before it may be tested. 0. The commands below must run as root, so log in as root, or use a program such as "su" to become root temporarily. 1. Create a datapath instance. The command below creates a datapath with ID 0 (see dpctl(8) for more detailed usage information). # dpctl adddp 0 In principle, openflow_mod supports multiple datapaths within the same host, but this is rarely useful in practice. If you built a support module for hardware accelerated OpenFlow switching and you want to use it, you must load it before creating the datapath with "dpctl adddp". 2. Use dpctl to attach the datapath to physical interfaces on the machine. Say, for example, you want to create a trivial 2-port switch using interfaces eth1 and eth2, you would issue the following commands: # dpctl addif 0 eth1 # dpctl addif 0 eth2 You can verify that the interfaces were successfully added by asking dpctl to print the current status of datapath 0: # dpctl show 0 3. (Optional) You can manually add flows to the datapath to test using dpctl add-flows and view them using dpctl dump-flows. See dpctl(8) for more details. 4. The simplest way to test the datapath is to run the provided sample controller on the host machine to manage the datapath directly using netlink: # controller -v nl:0 Once the controller is running, the datapath should operate like a learning Ethernet switch. You may monitor the flows in the datapath flow table using "dpctl dump-flows" command. The preceding instructions assume that the controller and the switch are running on the same machine. This is an easy configuration for testing, but a more conventional setup would run a controller on one machine and one or more switches on different machines. Use the following instructions to set up remote switches: 1. Start the datapath and attach it to two or more physical ports as described in the previous section. 2. Run the controller in passive TCP mode on the host which will act as the controller. In the example below, the controller will bind to port 975 (the default) awaiting connections from secure channels. # controller -v ptcp: (See controller(8) for more details) Make sure the machine hosting the controller is reachable by the switch. 3. Arrange so that the switch can reach the controller over the network. There are two ways to do this: - Use a "control network" that is completely separate from the "data network" to be controlled. To do so, configure a network device (one that has not been added to the datapath with "dpctl addif") to access the control network in the usual way. - Use the same network for control and for data. For this purpose, each datapath nl:K has a corresponding virtual network device named ofK. Start by bringing up of0 before you start the secure channel: # ifconfig of0 up Before the secure channel starts up, the of0 device cannot send or receive any packets, so the next step depends on whether connectivity is required to configure the device's IP address: . If the switch has a static IP address, you may configure its IP address now, e.g.: # ifconfig of0 192.168.1.1 . If the switch does not have a static IP address, e.g. its IP address is obtained dynamically via DHCP, then proceed to step 4. The DHCP client will not be able to contact the DHCP server until the secure channel has started up. 4. Run secchan on the datapath host to start the secure channel connecting the datapath to a remote controller. (See secchan(8) for usage details). The channel should be configured to connect to the controller's IP address on the port configured in step 2. If the controller is running on host 192.168.1.2 port 975 (the default port) and the datapath ID is 0, the secchan invocation would look like: # secchan -v nl:0 tcp:192.168.1.2 If you are using separate control and data networks, or if the networks are combined and the switch has a static IP address, the secure channel should quickly connect to the controller. Setup is now complete. Otherwise, proceed to step 5. 5. If you are using the same network for control and data, and the switch obtains its IP address dynamically, then you may now obtain the switch's IP address, e.g. by invoking a DHCP client. The secure channel will only be able to connect to the controller after an IP address has been obtained. Bug Reporting ------------- Please report problems to: info@openflowswitch.org