From: Justin Pettit Date: Tue, 8 Sep 2009 03:50:06 +0000 (-0700) Subject: in-band: Document logic behind in-band's design. X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=850887473a7093066db80569c9f2f9097941d6bc;p=openvswitch in-band: Document logic behind in-band's design. There have been numerous attempts at getting in-band correct. If history is at all an example, it probably still isn't. However, this is an attempt to document its current design, so that we can understand what our current thinking is. --- diff --git a/secchan/in-band.c b/secchan/in-band.c index fa77a254..a2197c6a 100644 --- a/secchan/in-band.c +++ b/secchan/in-band.c @@ -43,18 +43,168 @@ #define THIS_MODULE VLM_in_band #include "vlog.h" +/* In-band control allows a single network to be used for OpenFlow + * traffic and other data traffic. Refer to ovs-vswitchd.conf(5) and + * secchan(8) for a description of configuring in-band control. + * + * This comment is an attempt to describe how in-band control works at a + * wire- and implementation-level. Correctly implementing in-band + * control has proven difficult due to its many subtleties, and has thus + * gone through many iterations. Please read through and understand the + * reasoning behind the chosen rules before making modifications. + * + * In Open vSwitch, in-band control is implemented as "hidden" flows (in + * that they are not visible through OpenFlow) and at a higher priority + * than wildcarded flows can be setup by the controller. This is done + * so that the controller cannot interfere with them and possibly break + * connectivity with its switches. It is possible to see all flows, + * including in-band ones, with the ovs-appctl "bridge/dump-flows" + * command. + * + * The following rules are always enabled with the "normal" action by a + * switch with in-band control: + * + * a. DHCP requests sent from the local port. + * b. ARP replies to the local port's MAC address. + * c. ARP requests from the local port's MAC address. + * d. ARP replies to the remote side's MAC address. Note that the + * remote side is either the controller or the gateway to reach + * the controller. + * e. ARP requests from the remote side's MAC address. Note that + * like (d), the MAC is either for the controller or gateway. + * f. ARP replies containing the controller's IP address as a target. + * g. ARP requests containing the controller's IP address as a source. + * h. OpenFlow (6633/tcp) traffic to the controller's IP. + * i. OpenFlow (6633/tcp) traffic from the controller's IP. + * + * The goal of these rules is to be as narrow as possible to allow a + * switch to join a network and be able to communicate with a + * controller. As mentioned earlier, these rules have higher priority + * than the controller's rules, so if they are too broad, they may + * prevent the controller from implementing its policy. As such, + * in-band actively monitors some aspects of flow and packet processing + * so that the rules can be made more precise. + * + * In-band control monitors attempts to add flows into the datapath that + * could interfere with its duties. The datapath only allows exact + * match entries, so in-band control is able to be very precise about + * the flows it prevents. Flows that miss in the datapath are sent to + * userspace to be processed, so preventing these flows from being + * cached in the "fast path" does not affect correctness. The only type + * of flow that is currently prevented is one that would prevent DHCP + * replies from being seen by the local port. For example, a rule that + * forwarded all DHCP traffic to the controller would not be allowed, + * but one that forwarded to all ports (including the local port) would. + * + * As mentioned earlier, packets that miss in the datapath are sent to + * the userspace for processing. The userspace has its own flow table, + * the "classifier", so in-band checks whether any special processing + * is needed before the classifier is consulted. If a packet is a DHCP + * response to a request from the local port, the packet is forwarded to + * the local port, regardless of the flow table. Note that this requires + * L7 processing of DHCP replies to determine whether the 'chaddr' field + * matches the MAC address of the local port. + * + * It is interesting to note that for an L3-based in-band control + * mechanism, the majority of rules are devoted to ARP traffic. At first + * glance, some of these rules appear redundant. However, each serves an + * important role. First, in order to determine the MAC address of the + * remote side (controller or gateway) for other ARP rules, we must allow + * ARP traffic for our local port with rules (b) and (c). If we are + * between a switch and its connection to the controller, we have to + * allow the other switch's ARP traffic to through. This is done with + * rules (d) and (e), since we do not know the addresses of the other + * switches a priori, but do know the controller's or gateway's. Finally, + * if the controller is running in a local guest VM that is not reached + * through the local port, the switch that is connected to the VM must + * allow ARP traffic based on the controller's IP address, since it will + * not know the MAC address of the local port that is sending the traffic + * or the MAC address of the controller in the guest VM. + * + * With a few notable exceptions below, in-band should work in most + * network setups. The following are considered "supported' in the + * current implementation: + * + * - Locally Connected. The switch and controller are on the same + * subnet. This uses rules (a), (b), (c), (h), and (i). + * + * - Reached through Gateway. The switch and controller are on + * different subnets and must go through a gateway. This uses + * rules (a), (b), (c), (h), and (i). + * + * - Between Switch and Controller. This switch is between another + * switch and the controller, and we want to allow the other + * switch's traffic through. This uses rules (d), (e), (h), and + * (i). It uses (b) and (c) indirectly in order to know the MAC + * address for rules (d) and (e). Note that DHCP for the other + * switch will not work unless the controller explicitly lets this + * switch pass the traffic. + * + * - Between Switch and Gateway. This switch is between another + * switch and the gateway, and we want to allow the other switch's + * traffic through. This uses the same rules and logic as the + * "Between Switch and Controller" configuration described earlier. + * + * - Controller on Local VM. The controller is a guest VM on the + * system running in-band control. This uses rules (a), (b), (c), + * (h), and (i). + * + * - Controller on Local VM with Different Networks. The controller + * is a guest VM on the system running in-band control, but the + * local port is not used to connect to the controller. For + * example, an IP address is configured on eth0 of the switch. The + * controller's VM is connected through eth1 of the switch, but an + * IP address has not been configured for that port on the switch. + * As such, the switch will use eth0 to connect to the controller, + * and eth1's rules about the local port will not work. In the + * example, the switch attached to eth0 would use rules (a), (b), + * (c), (h), and (i) on eth0. The switch attached to eth1 would use + * rules (f), (g), (h), and (i). + * + * The following are explicitly *not* supported by in-band control: + * + * - Specify Controller by Name. Currently, the controller must be + * identified by IP address. A naive approach would be to permit + * all DNS traffic. Unfortunately, this would prevent the + * controller from defining any policy over DNS. Since switches + * that are located behind us need to connect to the controller, + * in-band cannot simply add a rule that allows DNS traffic from + * the local port. The "correct" way to support this is to parse + * DNS requests to allow all traffic related to a request for the + * controller's name through. Due to the potential security + * problems and amount of processing, we decided to hold off for + * the time-being. + * + * - Multiple Controllers. There is nothing intrinsic in the high- + * level design that prevents using multiple (known) controllers, + * however, the current implementation's data structures assume + * only one. + * + * - Differing Controllers for Switches. All switches must know + * the L3 addresses for all the controllers that other switches + * may use, since rules need to be setup to allow traffic related + * to those controllers through. See rules (f), (g), (h), and (i). + * + * - Differing Routes for Switches. In order for the switch to + * allow other switches to connect to a controller through a + * gateway, it allows the gateway's traffic through with rules (d) + * and (e). If the routes to the controller differ for the two + * switches, we will not know the MAC address of the alternate + * gateway. + */ + #define IB_BASE_PRIORITY 18181800 enum { - IBR_FROM_LOCAL_DHCP, /* From local port, DHCP. */ - IBR_TO_LOCAL_ARP, /* To local port, ARP. */ - IBR_FROM_LOCAL_ARP, /* From local port, ARP. */ - IBR_TO_REMOTE_ARP, /* To remote MAC, ARP. */ - IBR_FROM_REMOTE_ARP, /* From remote MAC, ARP. */ - IBR_TO_CTL_ARP, /* To controller IP, ARP. */ - IBR_FROM_CTL_ARP, /* From controller IP, ARP. */ - IBR_TO_CTL_OFP, /* To controller, OpenFlow port. */ - IBR_FROM_CTL_OFP, /* From controller, OpenFlow port. */ + IBR_FROM_LOCAL_DHCP, /* (a) From local port, DHCP. */ + IBR_TO_LOCAL_ARP, /* (b) To local port, ARP. */ + IBR_FROM_LOCAL_ARP, /* (c) From local port, ARP. */ + IBR_TO_REMOTE_ARP, /* (d) To remote MAC, ARP. */ + IBR_FROM_REMOTE_ARP, /* (e) From remote MAC, ARP. */ + IBR_TO_CTL_ARP, /* (f) To controller IP, ARP. */ + IBR_FROM_CTL_ARP, /* (g) From controller IP, ARP. */ + IBR_TO_CTL_OFP, /* (h) To controller, OpenFlow port. */ + IBR_FROM_CTL_OFP, /* (i) From controller, OpenFlow port. */ #if OFP_TCP_PORT != OFP_SSL_PORT #error Need to support separate TCP and SSL flows. #endif