PCP working group M. Boucadair Internet-Draft France Telecom Intended status: Standards Track F. Dupont Expires: July 10, 2011 Internet Systems Consortium R. Penno Juniper Networks January 6, 2011 Port Control Protocol (PCP) Failure Scenarios draft-boucadair-pcp-failure-00 Abstract This document identifies and analyses several PCP failure scenarios. A procedure to retrieve the mapping(s) from the PCP Server is proposed. This procedure relies upon the use of two new PCP OpCodes: GET and GETNEXT. Requirements Language The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 [RFC2119]. Status of this Memo This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet- Drafts is at http://datatracker.ietf.org/drafts/current/. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress." This Internet-Draft will expire on July 10, 2011. Copyright Notice Copyright (c) 2011 IETF Trust and the persons identified as the document authors. All rights reserved. This document is subject to BCP 78 and the IETF Trust's Legal Boucadair, et al. Expires July 10, 2011 [Page 1] Internet-Draft PCP Failure Scenarios January 2011 Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License. Table of Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 2. PCP Failure Scenarios . . . . . . . . . . . . . . . . . . . . 3 2.1. Change of the IP Address of The PCP Server . . . . . . . . 3 2.2. Application Crash . . . . . . . . . . . . . . . . . . . . 3 2.3. PCP Client Crash . . . . . . . . . . . . . . . . . . . . . 4 2.4. Change of the Internal IP Address . . . . . . . . . . . . 4 2.5. Change of the CPE WAN IP Address . . . . . . . . . . . . . 5 2.6. Restart or Failure of the PCP Server . . . . . . . . . . . 5 2.6.1. Basic Rule . . . . . . . . . . . . . . . . . . . . . . 5 2.6.2. Clear PCP Mappings . . . . . . . . . . . . . . . . . . 6 2.6.3. State Redundancy is Enabled . . . . . . . . . . . . . 6 2.6.4. Cold-Standby without State Redundancy . . . . . . . . 6 2.6.5. Anycast Redundancy Mode . . . . . . . . . . . . . . . 6 3. PCP State Synchronization Overview . . . . . . . . . . . . . . 6 4. GET and GETNEXT OpCodes . . . . . . . . . . . . . . . . . . . 7 4.1. Message Format . . . . . . . . . . . . . . . . . . . . . . 7 4.2. New Result Codes . . . . . . . . . . . . . . . . . . . . . 8 4.3. GET: Retrieving One Single Mapping . . . . . . . . . . . . 9 4.4. GETNEXT: Retrieve the Mapping Table . . . . . . . . . . . 9 4.4.1. Filter . . . . . . . . . . . . . . . . . . . . . . . . 9 4.4.2. Mapping Table Order . . . . . . . . . . . . . . . . . 10 4.4.3. Behavior . . . . . . . . . . . . . . . . . . . . . . . 10 4.4.4. Misc . . . . . . . . . . . . . . . . . . . . . . . . . 11 4.5. Flow Examples . . . . . . . . . . . . . . . . . . . . . . 11 5. Security Considerations . . . . . . . . . . . . . . . . . . . 17 6. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 17 7. References . . . . . . . . . . . . . . . . . . . . . . . . . . 17 7.1. Normative References . . . . . . . . . . . . . . . . . . . 17 7.2. Informative References . . . . . . . . . . . . . . . . . . 17 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 18 Boucadair, et al. Expires July 10, 2011 [Page 2] Internet-Draft PCP Failure Scenarios January 2011 1. Introduction This document discusses several failure scenarios that may occurs when deploying PCP. The purpose is to kick off discussion about these concerns and to integrate a consolidated version into the base PCP specification document [I-D.ietf-pcp-base]. 2. PCP Failure Scenarios 2.1. Change of the IP Address of The PCP Server When a new IP address is used to reach its PCP Server, the PCP Client MUST re-create all of its mappings using the newly discovered IP address. The PCP Client MUST undertake the same process as per refreshing an existing mapping (see [I-D.ietf-pcp-base]); the only difference is the PCP Requests are sent to a distinct IP address. No specific behavior is required from the PCP Server for handling these requests. 2.2. Application Crash When a fatal error is encountered by an application relying on PCP to open pinholes on an upstream device, and upon the restart of that application, the PCP Client should issue appropriate requests to refresh the mappings of that application (e.g., clear old mappings and install new ones using the new port number used by the application). If a distinct port number is used by the application to bound its service (i.e., a new internal port number is to be signaled in PCP), the PCP Server may honor the refresh requests if the per-subscriber quota is not exceeded. A distinct external port number would be assigned by the PCP Server due to the presence of "stale" mapping(s) associated with the "old" port number. To avoid this inconvenience induced by stale mappings, the PCP Client MAY clear the "old" mappings before issuing the refresh requests; but this would require the PCP Client to store the information about the "old" port number. This may not be easy in some scenarios because the PCP Client may handle PCP requests on behalf of several applications and no means to identify the requesting application may be supported. Means to identify the application are implementation- specific and are out of scope of this document. Boucadair, et al. Expires July 10, 2011 [Page 3] Internet-Draft PCP Failure Scenarios January 2011 It is NOT RECOMMENDED for a PCP Client to issue a request to delete all the mappings associated with an internal IP address since other applications and PCP Client(s) may use the same internal IP address to instruct their pinholes in the PCP Server. [Ed. This is in fact about the "application ID" idea.] 2.3. PCP Client Crash The PCP Client may encounter a fatal error leading to its restart. In such case, the internal IP address and port numbers used by requesting applications are not impacted. Therefore, the PCP mappings as maintained by the PCP Server are accurate and there is no need to refresh them. On the PCP Client side, a new UDP port should be assigned to issue PCP requests. As a consequence, if outstanding requests have been sent to the PCP Server, the responses are likely to be lost. [Ed. This is true only for the first NAT in a chain (or of course for one NAT chains). For second and further NAT in a chain we have again a sub-case of the amnesic client issue.] If the PCP Client stores its mapping in a persistent memory, there is no need to retrieve the list of active mappings from the PCP Server. If several PCP Clients are co-located on the same host, related PCP mapping tables should be uniquely distinguished (e.g., a PCP Client does not delete pinholes instructed by another PCP Client.) If the PCP Client (or the application) is crashing, it should be allocating short PCP lifetimes until it is debugged and running properly. If it is never debugged and never running properly, it should continue to request short PCP lifetimes. [Ed. This is in the base-02 now.] 2.4. Change of the Internal IP Address When a new IP address is assigned to a host embedding a PCP Client, the PCP Client MUST install on the PCP Server all the mappings it manages, using the new assigned IP address as the internal IP address. The hinted external port number won't be assigned by the PCP Server since a "stale" mapping is already instantiated by the PCP Server (but it is associated with a distinct internal IP address). Depending on the configured policies, the PCP Server may honor all or part of the requests received from the PCP Client. Upon receipt of the response from the PCP Server, the PCP Client MUST update its Boucadair, et al. Expires July 10, 2011 [Page 4] Internet-Draft PCP Failure Scenarios January 2011 local PCP state with the new assigned port numbers and external IP address. [Ed. Note: Do we need to support means to clear stale mappings first? This may have an impact if the quota is exceed due to the presence of stale mappings.] A PCP Client may be used to manage pinholes on behalf of a third party (i.e., the PCP Client and the third party are not co-located on the same host). If a new internal IP address is assigned to that third party (e.g., webcam), the PCP Client SHOULD be instructed to delete the old mapping(s) and create new one(s) using the new assigned internal IP address. 2.5. Change of the CPE WAN IP Address The change of the IP address of the WAN interface of the CPE would have an impact on the accuracy of the mappings instantiated in the PCP Server: o For the DS-Lite case [I-D.ietf-softwire-dual-stack-lite]: if a new IPv6 address is used by the B4 element when encapsulating IPv4 packets in IPv6 ones, the mappings SHOULD be refreshed: If the PCP Client is embedded in the B4, the refresh operation is triggered by the change of the B4 IPv6 address. This would be more complicated when the PCP Client is located in a device behind the B4. [Ed. Note: how an IPv4 host behind a DS-Lite CPE is aware that a new IPv6 address is used by the B4?] o For the NAT64 case [I-D.ietf-behave-v6v4-xlate-stateful], any change of the assigned IPv6 prefix delegated to the CPE will be detected by the PCP Client (because this leads to the allocation of a new IPv6 address). The PCP Client has to undertake the operation described in Section 2.4. 2.6. Restart or Failure of the PCP Server This section covers failure scenarios encountered by the PCP Server. 2.6.1. Basic Rule In any situation the PCP Server loses all or part of its PCP state, the Epoch value MUST be reset when replying to received requests. Doing so would allow PCP Client to audit its mapping table. If the state is not lost, the PCP Server MUST NOT reset the Epoch Boucadair, et al. Expires July 10, 2011 [Page 5] Internet-Draft PCP Failure Scenarios January 2011 value returned to requesting PCP Clients. 2.6.2. Clear PCP Mappings When a command line or a configuration change is enforced to clear all or a subset of PCP mappings maintained by the PCP Server, the PCP Server MUST reset its Epoch to zero value. 2.6.3. State Redundancy is Enabled When state redundancy is enabled, the state is not lost during failure events. Failures are therefore transparent to requesting PCP Clients. When a backup device takes over, Epoch MUST NOT be reset to zero. 2.6.4. Cold-Standby without State Redundancy In this section we assume that a redundancy mechanisms is configured between a primary PCP-controlled device and a backup one but without activating any state synchronization for the PCP-instructed mappings between the backup and the primary devices. If the primary PCP-controlled device fails and the backup one takes over, the PCP Server MUST reset the Epoch to zero value. Doing so would allow PCP Clients to detect the loss of states in the PCP Server and proceed to state synchronization. 2.6.5. Anycast Redundancy Mode When an anycast-based mode is deployed (i.e., the same IP address is used to reach several PCP Servers) for redundancy reasons, the change of the PCP Server which handles the requests of a given PCP Client won't be detected by that PCP Client. Relying on the Epoch to detect the loss of state won't help in this scenario to re-create missing mappings. Proprietary solutions MAY be envisaged to coordinate amongst anycasted PCP Servers; otherwise the use of the anycast is NOT RECOMMENDED. 3. PCP State Synchronization Overview The following sketches the state synchronization logic: Boucadair, et al. Expires July 10, 2011 [Page 6] Internet-Draft PCP Failure Scenarios January 2011 o The PCP Client/host/application maintains state; o The PCP Server maintains state; o If the PCP Client (resp., the PCP Server) crashes and restarts it just have to synchronize with the PCP Server (resp., the PCP Client); o If both crash then one has to use stable storage and we fall back in the previous case as soon as we know which one (the Epoch value gives this information); o PCP Server -> PCP Client synchronization requires a GET mechanism to retrieve the state from the PCP Server; o PCP Client -> PCP Server synchronization is done by a re-create or refresh of the state. The PCP Client MAY retrieve the PCP Server state in order to prevent stale mappings. 4. GET and GETNEXT OpCodes This section defines two new PCP OpCodes called: GET and GETNEXT. These PCP methods are used by the PCP Client to retrieve a mapping or to walk through the mapping table maintained by the PCP Server and retrieves a list of mapping entries it instantiated. 4.1. Message Format The layout of GET and GETNEXT OpCodes request is shown in Figure 1. Figure 2 shows the message format of GET and GETNEXT response. Boucadair, et al. Expires July 10, 2011 [Page 7] Internet-Draft PCP Failure Scenarios January 2011 0 1 2 3 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Protocol | Reserved | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | : Pinhole Internal IP address (32 or 128, depending on OpCode) : | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | : Assigned external IP address (32 or 128, depending on OpCode) : | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | all zeros | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | internal port | assigned external port | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Figure 1: GET/GETNEXT: Message format of the request 0 1 2 3 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Protocol | Reserved | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | : Pinhole Internal IP address (32 or 128, depending on OpCode) : | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | : Assigned external IP address (32 or 128, depending on OpCode) : | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Remaining lifetime | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | internal port | assigned external port | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Figure 2: GET/GETNEXT: Message format of the response 4.2. New Result Codes TBD - NO_MAPPING_FOUND, e.g., there is no matching mapping. TBD - NO_MORE_MAPPING, e.g., there is no greater matching mapping. Boucadair, et al. Expires July 10, 2011 [Page 8] Internet-Draft PCP Failure Scenarios January 2011 Figure 3: GET/GETNEXT-Specific Result Codes 4.3. GET: Retrieving One Single Mapping In order to retrieve a mapping with known attributes (e.g., the internal IP address, internal port number and the protocol), the PCP Client issues a GET request having these attributes set to their known values, and all others fields set to wildcard values (i.e., zero). The lifetime field SHOULD be set to zero and MUST be ignored on reception. Upon receipt of the request, the PCP Server checks whether a mapping matching the given attributes exists. If one mapping is found a GET response enclosing the mapping information is returned to the PCP Client. If no mapping is found, the PCP Server returns back a GET response with result code set to NO_MAPPING_FOUND. If more than one mapping is found, the PCP Server returns back a GET response with result code set to AMBIGUOUS, and the PCP Client can retry with a more specific GET request. [Ed. Two sets of attributes characterize a mapping (they are used to lookup a mapping with an incoming or outgoing packet): the internal set (internal address and port, and protocol) and the external set (external address and port, and protocol). If End- point Independent Mode is not used, the remote peer address and port (i.e., REMOTE_PEER) can be needed too. When a PCP GET response indicating an error code equal to AMBIGUOUS is received, the PCP Client may adopt two strategies: The REMOTE_PEER option can be used to narrow the lockup; the PCP Client issues a GET request with a REMOTE_PEER filter set to a valid IP address (and port); or the PCP Client issues a GETNEXT request having all known attributes set (i.e., the internal IP address, internal port number and the protocol). The PCP Client and the PCP Server undertake the procedure specified in Section 4.4 to retrieve all the mappings having the same (internal IP address, internal port number and the protocol). REMOTE_PEER option can be used to retrieve its value.] 4.4. GETNEXT: Retrieve the Mapping Table The GETNEXT operation allows a PCP Client to incrementally retrieve the an image of the PCP Server's mapping table. 4.4.1. Filter A filter denotes a set of attributes indicated by the PCP Client to a PCP Server in a GETNEXT request. Only mappings matching that filter Boucadair, et al. Expires July 10, 2011 [Page 9] Internet-Draft PCP Failure Scenarios January 2011 will be returned to the requesting PCP Client. Currently the only defined filter is the value of the internal IP address field; e.g.- with a not wildcard (i.e., not zero) internal IP address in GETNEXT requests, the PCP Server should return all the mappings having an internal IP address equal to this address. 4.4.2. Mapping Table Order A PCP Server MUST define a fixed order between mappings with the all zero mapping as the minimum element. It MAY use a lexicographical order on mapping attributes. [Ed. To prevent the removal of a pinhole in the PCP Server due to a PCP Client failure, mappings with lower remaining lifetime MAY be returned first.] [Ed. The problem is this order is not stable: a mapping deletion or renewal will change the order and mess a slow retrieval.] 4.4.3. Behavior To retrieve the mapping table, GETNEXT is used by a PCP Client to retrieve the first mapping from the PCP Server matching the filter (i.e., internal IP address). When issuing the GETNEXT request, the internal IP address is set to a valid IP address or the wildcard (zero). All the remaining fields SHOULD be set to zero. The PCP Server returns the first mapping matching the filter indicated by the PCP Client. The PCP Server indicates in the GETNEXT response all the information related to the mapping. Upon receipt of the response, the PCP Client retrieves the enclosed mapping and stores it in its mapping table image. Then GETNEXT request is sent by the PCP Client. All the fields of the GETNEXT request except the filtering fields and the remaining lifetime are copied from the received GETNEXT response from the PCP Server. When GETNEXT request is received by the PCP Server, it reads the enclosed mapping and proceeds to a lookup of its mapping table to find the mapping having the attributes indicated in the GETNEXT request. If none is found it returns the first matching mapping, if more than one is found it returns back a GETNEXT response with result code set to AMBIGUOUS. If one is found, it returns either the next matching mapping when it exists, or a GETNEXT response with result code set to NO_MORE_MAPPING. This procedure is repeated by the PCP Client until retrieving all its Boucadair, et al. Expires July 10, 2011 [Page 10] Internet-Draft PCP Failure Scenarios January 2011 mappings from the PCP Server. Termination is indicated by NO_MORE_MAPPING. The REMOTE_PEER option can be used to get its value. Two strategies are possible for the PCP client: either it includes a REMOTE_PEER option in all GETNEXT requests, or it waits for an AMBIGUOUS error and uses a REMOTE_PEER option to solve conflicts. [Ed. Wildcard REMOTE_PEER value is not defined in the pcp-base document.] [Ed. Wildcard protocol to indicate ANY protocol is not defined in the pcp-base document.] 4.4.4. Misc This section will be completed latter: * lock or not the mapping table in the server side? * with locks, when to unlock? * postpone PINxy requests until the synchronization process is completed? 4.5. Flow Examples As an illustration example, let's consider the following mapping table is maintained by the PCP Server: +-----+--------------+----------+-----------+----------+------------+ | Pro | Internal IP | Internal | External | External | Remaining | | | Address | Port | IP | Port | Lifetime | | | | | Address | | | +-----+--------------+----------+-----------+----------+------------+ | UDP | 198.51.100.1 | 25655 | 192.0.2.1 | 15659 | 1659 | | TCP | 198.51.100.2 | 12354 | 192.0.2.1 | 32654 | 3600 | | TCP | 198.51.100.2 | 8596 | 192.0.2.1 | 25659 | 6000 | | UDP | 198.51.100.1 | 19856 | 192.0.2.1 | 42654 | 7200 | | TCP | 198.51.100.1 | 15775 | 192.0.2.1 | 32652 | 9000 | +-----+--------------+----------+-----------+----------+------------+ Table 1: Excerpt of a mapping table As shown in Table 1, the PCP Server sorts the mapping table using the internal IP address and the remaining lifetime. Figure 5 illustrates the exchange that occurs when a PCP Client tries Boucadair, et al. Expires July 10, 2011 [Page 11] Internet-Draft PCP Failure Scenarios January 2011 to retrieve the information related to a non-existing mapping. Figure 5 shows an example of a PCP Client which retrieves successfully an existing mapping from the PCP Server. +------+ +------+ | PCP | | PCP | |Client| |Server| +------+ +------+ | (1) PCP GET Request | | protocol= TCP | | internal-ip-address= 198.51.100.1 | | internal-port= 59864 | |---------------------------------->| | | | (2) PCP GET Response | | error= NO_MAPPING_FOUND | |<----------------------------------| | | Figure 4: Example of a failed GET operation Figure 5 shows an example of a PCP Client which retrieves successfully an existing mapping from the PCP Server. Boucadair, et al. Expires July 10, 2011 [Page 12] Internet-Draft PCP Failure Scenarios January 2011 +------+ +------+ | PCP | | PCP | |Client| |Server| +------+ +------+ | (1) PCP GET Request | | protocol= TCP | | internal-ip-address= 198.51.100.1 | | internal-port= 25655 | |---------------------------------->| | | | (2) PCP GET Response | | protocol= TCP | | internal-ip-address= 198.51.100.1 | | internal-port= 25655 | | external-ip-address= 192.0.2.1 | | external-port= 15659 | | remaining-lifetime= 1659 | |<----------------------------------| | | | (3) PCP PIN44 Request | | protocol= TCP | | internal-ip-address= 198.51.100.1 | | internal-port= 25655 | | external-ip-address= 192.0.2.1 | | external-port= 15659 | | requested-lifetime= 0 | |---------------------------------->| | | Figure 5: Example of a successful GET operation In reference to Figure 6, the PCP Server returns first the mapping having the internal address equal to 192.0.2.1 with the lower remaining lifetime. Boucadair, et al. Expires July 10, 2011 [Page 13] Internet-Draft PCP Failure Scenarios January 2011 +------+ +------+ | PCP | | PCP | |Client| |Server| +------+ +------+ | (1) PCP GETNEXT Request | | internal-ip-address= 198.51.100.2 | |---------------------------------->| | | | (2) PCP GETNEXT Response | | protocol= TCP | | internal-ip-address= 198.51.100.2 | | internal-port= 12354 | | external-ip-address= 192.0.2.1 | | external-port= 32654 | | remaining-lifetime= 3600 | |<----------------------------------| | | | (3) PCP GETNEXT Request | | protocol= TCP | | internal-ip-address= 198.51.100.2 | | internal-port= 12354 | | external-ip-address= 192.0.2.1 | | external-port= 32654 | |---------------------------------->| | | | (4) PCP GETNEXT Response | | protocol= TCP | | internal-ip-address= 198.51.100.2 | | internal-port= 8596 | | external-ip-address= 192.0.2.1 | | external-port= 25659 | | remaining-lifetime= 6000 | |<----------------------------------| | | | (5) PCP GETNEXT Request | | protocol= TCP | | internal-ip-address= 198.51.100.2 | | internal-port= 8596 | | external-ip-address= 192.0.2.1 | | external-port= 25659 | |---------------------------------->| | | | (6) PCP GETNEXT Response | | error= NO_MORE_MAPPING | |<----------------------------------| | | Figure 6: Flow example of GETNEXT Boucadair, et al. Expires July 10, 2011 [Page 14] Internet-Draft PCP Failure Scenarios January 2011 In reference to Figure 7, the PCP Server returns first the mapping having the internal address equal to 192.0.2.2 with the lower remaining lifetime. In this example, the same internal port is used for TCP and UDP. Boucadair, et al. Expires July 10, 2011 [Page 15] Internet-Draft PCP Failure Scenarios January 2011 +------+ +------+ | PCP | | PCP | |Client| |Server| +------+ +------+ | (1) PCP GETNEXT Request | | internal-ip-address= 198.51.100.1 | |---------------------------------->| | | | (2) PCP GETNEXT Response | | protocol= UDP | | internal-ip-address= 198.51.100.1 | | internal-port= 25655 | | external-ip-address= 192.0.2.1 | | external-port= 15659 | | remaining-lifetime= 1659 | |<----------------------------------| | | | (3) PCP GETNEXT Request | | protocol= UDP | | internal-ip-address= 198.51.100.1 | | internal-port= 25655 | | external-ip-address= 192.0.2.1 | | external-port= 15659 | |---------------------------------->| | | | (4) PCP GETNEXT Response | | protocol= TCP | | internal-ip-address= 198.51.100.1 | | internal-port= 25655 | | external-ip-address= 192.0.2.1 | | external-port= 32652 | | remaining-lifetime= 9000 | |<----------------------------------| | | | (5) PCP GETNEXT Request | | protocol= TCP | | internal-ip-address= 198.51.100.1 | | internal-port= 25655 | | external-ip-address= 192.0.2.1 | | external-port= 32652 | |---------------------------------->| | | | (6) PCP GETNEXT Response | | error= NO_MORE_MAPPING | |<----------------------------------| | | Figure 7: Flow example of GETNEXT: same internal port number Boucadair, et al. Expires July 10, 2011 [Page 16] Internet-Draft PCP Failure Scenarios January 2011 5. Security Considerations TBD. [Ed. Two comments: * About the stable storage if this scenario is possible: 1. subscriber A gets a mapping 2. the PCP Server crashes and reboots 3. subscriber B gets the same mapping then the PCP Server MUST keep its state in a stable storage, i.e., it MUST NOT forget mappings. * About GET/GETNEXT, typically if a PCP Client is allowed to delete a mapping it SHOULD be allowed to retrieve it; and if it is not allowed to delete a mapping it MUST NOT be allowed to retrieve it.] 6. IANA Considerations TBD. 7. References 7.1. Normative References [I-D.ietf-pcp-base] Wing, D., "Port Control Protocol (PCP)", draft-ietf-pcp-base-02 (work in progress), January 2011. [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. 7.2. Informative References [I-D.ietf-behave-v6v4-xlate-stateful] Bagnulo, M., Matthews, P., and I. Beijnum, "Stateful NAT64: Network Address and Protocol Translation from IPv6 Clients to IPv4 Servers", draft-ietf-behave-v6v4-xlate-stateful-12 (work in progress), July 2010. Boucadair, et al. Expires July 10, 2011 [Page 17] Internet-Draft PCP Failure Scenarios January 2011 [I-D.ietf-softwire-dual-stack-lite] Durand, A., Droms, R., Woodyatt, J., and Y. Lee, "Dual- Stack Lite Broadband Deployments Following IPv4 Exhaustion", draft-ietf-softwire-dual-stack-lite-06 (work in progress), August 2010. Authors' Addresses Mohamed Boucadair France Telecom Rennes, 35000 France Email: mohamed.boucadair@orange-ftgroup.com Francis Dupont Internet Systems Consortium Email: fdupont@isc.org Reinaldo Penno Juniper Networks 1194 N Mathilda Avenue Sunnyvale, California 94089 USA Email: rpenno@juniper.net Boucadair, et al. Expires July 10, 2011 [Page 18]