idnits 2.17.1 draft-ietf-opsec-ns-impact-02.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year == Line 69 has weird spacing: '...ecurity and S...' == The document doesn't use any RFC 2119 keywords, yet seems to have RFC 2119 boilerplate text. -- The document date (July 28, 2020) is 1368 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- ** Obsolete normative reference: RFC 5246 (Obsoleted by RFC 8446) ** Obsolete normative reference: RFC 7525 (Obsoleted by RFC 9325) == Outdated reference: A later version (-18) exists of draft-ietf-tls-esni-07 Summary: 2 errors (**), 0 flaws (~~), 4 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 OPSEC Working Group N. Cam-Winget 3 Internet-Draft E. Wang 4 Intended status: Informational Cisco Systems, Inc. 5 Expires: January 29, 2021 R. Danyliw 6 Software Engineering Institute 7 R. DuToit 8 Broadcom 9 July 28, 2020 11 Impact of TLS 1.3 to Operational Network Security Practices 12 draft-ietf-opsec-ns-impact-02 14 Abstract 16 Network-based security solutions are used by enterprises, the public 17 sector, internet-service providers, and cloud-service providers to 18 both complement and enhance host-based security solutions. As TLS is 19 a widely deployed protocol to secure communication, these network- 20 based security solutions must necessarily interact with it. This 21 document describes this interaction for current operational security 22 practices and notes the impact of TLS 1.3 on them. 24 Status of This Memo 26 This Internet-Draft is submitted in full conformance with the 27 provisions of BCP 78 and BCP 79. 29 Internet-Drafts are working documents of the Internet Engineering 30 Task Force (IETF). Note that other groups may also distribute 31 working documents as Internet-Drafts. The list of current Internet- 32 Drafts is at https://datatracker.ietf.org/drafts/current/. 34 Internet-Drafts are draft documents valid for a maximum of six months 35 and may be updated, replaced, or obsoleted by other documents at any 36 time. It is inappropriate to use Internet-Drafts as reference 37 material or to cite them other than as "work in progress." 39 This Internet-Draft will expire on January 29, 2021. 41 Copyright Notice 43 Copyright (c) 2020 IETF Trust and the persons identified as the 44 document authors. All rights reserved. 46 This document is subject to BCP 78 and the IETF Trust's Legal 47 Provisions Relating to IETF Documents 48 (https://trustee.ietf.org/license-info) in effect on the date of 49 publication of this document. Please review these documents 50 carefully, as they describe your rights and restrictions with respect 51 to this document. Code Components extracted from this document must 52 include Simplified BSD License text as described in Section 4.e of 53 the Trust Legal Provisions and are provided without warranty as 54 described in the Simplified BSD License. 56 Table of Contents 58 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 59 2. Conventions and Definitions . . . . . . . . . . . . . . . . . 3 60 3. How TLS is used to enable Network-Based Security Solutions . 4 61 4. Changes in TLS 1.3 Relevant to Security Operations . . . . . 5 62 4.1. Perfect Forward Secrecy (PFS) . . . . . . . . . . . . . . 5 63 4.2. Encrypted Server Certificate . . . . . . . . . . . . . . 5 64 5. Network Security Operational Practices . . . . . . . . . . . 6 65 5.1. Passive TLS Inspection . . . . . . . . . . . . . . . . . 6 66 5.1.1. OP-1. Acceptable Use Policy (AUP) Enforcement (via 67 header inspection). . . . . . . . . . . . . . . . . . 7 68 5.1.2. OP-2. Network Behavior Analytics . . . . . . . . . . 7 69 5.1.3. OP-3. Crypto, Security and Security Policy 70 Compliance (server) . . . . . . . . . . . . . . . . . 8 71 5.1.4. OP-4. Crypto and Security Policy Compliance (client) 8 72 5.2. Outbound TLS Proxy . . . . . . . . . . . . . . . . . . . 9 73 5.2.1. OP-5: Acceptable Use Policy (AUP) Enforcement (via 74 payload inspection) . . . . . . . . . . . . . . . . . 10 75 5.2.2. OP-6: Data Loss Prevention Compliance . . . . . . . . 10 76 5.2.3. OP-7: Granular Network Segmentation . . . . . . . . . 10 77 5.2.4. OP-8: Network-based Threat Protection (client) . . . 10 78 5.2.5. OP-9: Protecting Challenging End Points . . . . . . . 11 79 5.2.6. OP-10: Content Injection . . . . . . . . . . . . . . 11 80 5.3. Inbound TLS Proxy . . . . . . . . . . . . . . . . . . . . 11 81 5.3.1. OP-11: TLS offloading . . . . . . . . . . . . . . . . 12 82 5.3.2. OP-12. Content distribution and application load 83 balancing . . . . . . . . . . . . . . . . . . . . . . 13 84 5.3.3. OP-13: Network-based Threat Protection (server) . . . 13 85 5.3.4. OP-14: Full Packet Capture . . . . . . . . . . . . . 13 86 5.3.5. OP-15: Application Layer Gateway (ALG) . . . . . . . 14 87 6. Security Considerations . . . . . . . . . . . . . . . . . . . 14 88 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 14 89 8. Appendix A: Summary Impact to Operational Practices with TLS 90 1.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 91 9. References . . . . . . . . . . . . . . . . . . . . . . . . . 15 92 9.1. Normative References . . . . . . . . . . . . . . . . . . 15 93 9.2. Informative References . . . . . . . . . . . . . . . . . 16 94 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . 17 95 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 17 97 1. Introduction 99 Enterprises, public sector organizations, internet service providers 100 and cloud service providers defend their networks and information 101 systems from attacks that originate from inside and outside their 102 networks. These organizations commonly employ security architectures 103 that involve complementary technologies deployed on both endpoints 104 and in the network; and collaborative watch-and-warning practices to 105 realize this defense. 107 The design of these security architectures and associated practices 108 entails numerous trade-offs. Typically, there is more than one 109 technical approach to realize a particular mitigation, although 110 comparable approaches may have different costs or side-effects. 111 Network-based solutions are often attractive to network 112 administrators because a single network device can: 114 o provide protection to many hosts and systems at once 116 o protect systems regardless of their type (e.g., fully patched 117 desktop systems on a modern operating system; unpatched function- 118 specific industrial control system) 120 o enforce policy on a system even if it is compromised, 121 misconfigured, not under configuration control or had its endpoint 122 protection disabled 124 o be managed (e.g. updates) and provisioned with resources (e.g. 125 disk and computing) independent of the systems it is protecting 127 o by itself, a single system may not be able to detect and mitigate 128 threats 130 In response to the adoption of new technologies, protocols and 131 threats, these security architectures must evolve to remain 132 effective. [RFC8404] documented a need to evolve with the effect of 133 pervasive encryption on operations. This document takes a narrower 134 focus by documenting the interaction of existing network-based 135 security practices with TLS 1.2 [RFC5246] (and earlier) traffic to 136 implement security policy, detection or mitigation of threats; and 137 the impact on these practices with improvements made in TLS 1.3 138 [RFC8446]. 140 2. Conventions and Definitions 142 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 143 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 144 "OPTIONAL" in this document are to be interpreted as described in BCP 145 14 [RFC2119] [RFC8174] when, and only when, they appear in all 146 capitals, as shown here. 148 Specific operational practices are numbered as "OP-##", operational 149 practice 1 (i.e., OP-1), 2 (i.e., OP-2), etc. 151 3. How TLS is used to enable Network-Based Security Solutions 153 Network-based security solutions come in many forms, most commonly as 154 Firewalls, Web Proxies, Intrusion Detection Systems (IDS), Intrusion 155 Prevention Systems (IPS) and Network Security Visibility and 156 Analytics systems. They inspect the network traffic, and then based 157 on their function, log their observation and/or act on the traffic to 158 implement security policy. When these devices act on the network 159 traffic, they are typically deployed inline as middleboxes (e.g. 160 firewalls) or as explicit proxies (e.g. web proxies). If their 161 function is only to observe, they can be deployed either as 162 middleboxes or given access to the network traffic out-of-band (OOB), 163 through the network fabric (e.g., network tap or span port). 165 Depending on their function, network-based security devices use 166 different degrees of visibility into the TLS traffic. Some 167 operational practices require only access to the unencrypted protocol 168 headers and associated meta-data of the TLS traffic. Other practices 169 require full visibility into the encrypted session (payload). 171 The practices that inspect only the unencrypted headers and meta-data 172 of TLS, require no special capabilities beyond access to the TLS 173 packets. However, to inspect the encrypted payload of TLS traffic 174 requires a TLS proxy. 176 A TLS proxy provides visibility and inspection to effectuate security 177 controls without changing the state machine of the TLS Server and TLS 178 Client, or the user experience. The TLS Proxy operates as a 179 transparent hop at the TLS layer in both middlebox and explicit proxy 180 deployments. For the web proxy case, after the client sends an HTTP 181 CONNECT to request a tunnel to the server, the web proxy may insert a 182 TLS Proxy function to proxy the TLS session without awareness by the 183 client or server. The TLS operation afterwards remains the same as a 184 middlebox. 186 To proxy a TLS session, a TLS Proxy must be able to present a valid 187 X.509 certificate to the TLS client to appear as a valid TLS Server; 188 similarly, the client must be able to validate the X.509 certificate 189 using the appropriate trust anchor for that TLS connection. To 190 achieve this, a deployment must properly provision their systems (TLS 191 Proxies and TLS clients). A TLS Proxy is unable to proxy a PSK based 192 session unless it is on-path and has proxied the session leading to 193 the PSK. TLS client authentication requires additional provisioning 194 for X.509 certificate on the TLS Server side. It does not have 195 impact on the deployment scenarios though. 197 4. Changes in TLS 1.3 Relevant to Security Operations 199 TLS 1.3 introduces a number of protocol design changes to improve 200 security and privacy. However, these enhancements impact current 201 network security operational practices that rely on the protocol 202 behavior of earlier TLS versions. 204 4.1. Perfect Forward Secrecy (PFS) 206 TLS 1.2 (and earlier versions) supports static RSA and Diffie-Hellman 207 (DH) cipher suites, which enables the server's private key to be 208 shared with a TLS proxy. [RFC7525] initiated the recommendation of 209 using AEAD cipher suites and specifically decoupling the cipher suite 210 negotiation based on the RSA key transport; this followed with TLS 211 1.3 explicitly removing support for these cipher suites in favor of 212 supporting only ephemeral mode Diffie-Hellman to provide perfect 213 forward secrecy (PFS). As a result of this enhancement, it would no 214 longer be possible for a server to share a key with the middlebox in 215 advance, which in turn implies that the middlebox cannot gain access 216 to the TLS session data.ss 218 4.2. Encrypted Server Certificate 220 TLS 1.2 (and earlier versions) sends the ClientHello, ServerHello and 221 Certificate messages in clear-text. In TLS 1.3, the Certificate 222 message is encrypted whereby hiding the server identity from any 223 intermediary. As a result of this enhancement, it would no longer be 224 possible to observe the server certificate without inspection the 225 encrypted TLS payload. 227 TLS proxies which implement a selective decryption policy will need 228 to alter their behavior to accommodate TLS 1.3. In TLS 1.2 (and 229 earlier), the proxy could observe the TLS handshake till seeing the 230 clear text server certificate to make the decryption policy decision. 231 For example, a proxy may not be permitted to decrypt certain types of 232 traffic such as those going to a banking and health care service. 233 However, in TLS 1.3, the TLS proxy must participate in both 234 handshakes (i.e., client-to-proxy; and proxy-to-server) in order to 235 view the server certificate. This change will impose a slight 236 increase in load per connection on the proxy. 238 5. Network Security Operational Practices 240 Specific network security operational practices applied to TLS 1.2 241 (and earlier) are described in subsequent sub-sections. They are 242 categorized into the following deployment scenarios: 244 1. Passive TLS inspection, where the network-based security function 245 is inspecting either the inbound or outbound TLS header or meta- 246 data traffic 248 2. Outbound TLS Proxy, where a TLS proxy mediates a TLS session 249 originating from a client inside the enterprise administrative 250 domain (and in the same administrative domain as the proxy) 251 towards an entity on the outside 253 3. Inbound TLS Proxy, where a TLS proxy mediates a TLS session from 254 a client outside the enterprise administrative domain towards an 255 entity on the inside (and in the same administrative domain as 256 the proxy) 258 Each deployment scenario describes current operational practices. 259 For each operational practice, possible deployment modes (e.g., 260 inline, out-of-band), a description of the practice, and the impact 261 of TLS 1.3 is categorized and explained. The categorized impacts to 262 practices when migrating to TLS 1.3 are as follows: 264 o no impact - no change in capability or performance is expected 265 with this practice 267 o no capability impact - no change in capability is expected; but 268 there may be a performance or implementation change required for 269 this practice 271 o reduced effectiveness - this practice will not be as effective on 272 TLS 1.3 traffic 274 o alternative approach required - this practice will not work with 275 TLS 1.3 traffic 277 It should be noted that [ECH] will further reduce the effectiveness 278 (passive inspection) or prevent certain practices (outbound proxy) 279 from being deployed. More study is required in this area. 281 5.1. Passive TLS Inspection 283 Passive TLS inspection is the deployment scenario where a network 284 security device passively inspects inbound or outbound TLS traffic to 285 make visibility inferences or take policy actions. The network 286 security device examines only the unencrypted TLS protocol headers 287 and does not have access to the encrypted content of the payload. 289 The TLS proxy deployment scenarios may also incorporate these 290 practices. 292 5.1.1. OP-1. Acceptable Use Policy (AUP) Enforcement (via header 293 inspection). 295 Deployment mode: inline 297 A firewall or web proxy restricts a client in the same administrative 298 domain from accessing sites or services outside that domain per an 299 acceptable use policy. The identification of the destination server 300 is performed through the inspection of either the SNI field in the 301 TLS ClientHello message from the client; or by extracting the server 302 identity from the Common Name (CN) or Subject Alternative Name (SAN) 303 fields of an X.509 certificate that is presented in the server's 304 Certificate TLS message. This data is used for domain categorization 305 or application identification. 307 This meta-data can also inform decryption eligibility decisions by a 308 firewall, in OP-4. For instance, a firewall may bypass traffic 309 decryption for a connection destined to a healthcare web service due 310 to privacy compliance requirements. 312 TLS 1.3 considerations: reduced effectiveness. Per Section 4.2, 313 domain categorization and application identification will be limited 314 to IP address and SNI information (beyond additional correlation 315 possible with other means such as DNS). 317 While an SNI is mandatory in TLS 1.3, there is no guarantee that the 318 server responding is the one indicated in the SNI from the client. A 319 SNI alone, without comparison of the server certificate, does not 320 provide reliable information about the server that the client is 321 attempting to reach. Where a client has been compromised by malware, 322 it may present an innocuous SNI to bypass protective filters (e.g., 323 to reach a command and control server), and this will be undetectable 324 under TLS 1.3. 326 5.1.2. OP-2. Network Behavior Analytics 328 Deployment mode: inline and out-of-band 330 Network behavior analysis and machine learning engines in IDSs, IPSs 331 and firewalls observe the cleartext fields of the TLS handshake 332 (e.g., session cipher suites) and conducts traffic analysis by 333 observing encrypted record sizes, packet rates and their inter- 334 arrival times, and similar outer connection behavior. They match 335 encrypted connections against known application patterns; identify 336 anomalies; and identify or block those without payload inspection. 337 These analytics may also observe that malicious applications may 338 deliberately manipulate certain TLS header fields, throttle packet 339 rates, and vary payload sizes in order to circumvent detection. 341 Through traffic analysis, researchers have detected devastating 342 pseudo-random number generator failures [TLS_VULNERABILITY], nonce 343 failures [NONCE_FAIL], and deeply flawed random number generators in 344 products in [WEAK_KEY] and [WEAK_K2]. 346 TLS 1.3 considerations: reduced effectiveness. Per Section 4.2, any 347 features relying on Certificate information will not be available. 349 5.1.3. OP-3. Crypto, Security and Security Policy Compliance (server) 351 Deployment: out-of-band 353 A network security device observes TLS handshake traffic to audit 354 that TLS server configuration conforms to policy. This compliance 355 monitoring commonly examines ciphersuites (e.g., use of weak 356 ciphersuites) and certificate properties (e.g., no self-signed 357 certificates, black or white list of certificate authorities, 358 certificate expiration times). 360 TLS 1.3 considerations: reduced effectiveness. Per Section 4.2, only 361 TLS ClientHello and ServerHello parameters can be audited. 362 Certification information will not be visible. 364 5.1.4. OP-4. Crypto and Security Policy Compliance (client) 366 Deployment: inline 368 A network security device observes TLS handshake traffic to ensure 369 that clients negotiating TLS connections have configurations (e.g., 370 only make connections with TLS 1.2+) and server certificate (e.g., 371 black-listed CAs) that adhere to policy. This is a variant of OP-3. 372 It is commonly used in deployments where an organization may have 373 reduced configuration control of end points (e.g., lab environments, 374 Bring Your Own Device arrangements, and IoT). 376 TLS 1.3 considerations: reduced effectiveness. Per Section 4.2, only 377 TLS ClientHello and ServerHello parameters can be audited. 378 Certification information will not be visible. 380 5.2. Outbound TLS Proxy 382 Outbound TLS proxy is the deployment scenario where a security device 383 that performs the TLS proxy function is in the same administrative 384 domain as the TLS client, and the TLS server is located in an 385 external zone such as the Internet or in another policy zone of the 386 same administrative domain. Usually the goal is to protect the 387 client endpoint and the organization by controlling application 388 behaviors and enforcing an acceptable use policy for the 389 organizational network. See Figure 1. 391 The administrator manages the TLS client to allow interception by the 392 TLS proxy, usually by deploying a local Certificate Authority (CA) 393 certificate on the TLS client. A typical scenario is an 394 organization-managed client endpoint, such as a laptop or a mobile 395 device that accesses the Internet through the organizational network. 396 When a client attempts to access an external TLS server, the TLS 397 proxy function typically presents a locally signed certificate from 398 the local CA on behalf of the server; alternatively, the certificate 399 generation function may be offloaded to an external Hardware Security 400 Module (HSM) service with which that the TLS proxy must integrate. 402 It has to be noted that the method does not work if the TLS client 403 does not support customized list of CAs, such as with certificate 404 pinning. The impact is independent of TLS 1.3 deployment. 406 _________ __________ 407 \ / 408 \ | Administrative 409 \ | Domain, _----__ 410 +-+ \ | Zone 2 / / \____ 411 | | \ \______/ __/ +------+ \ 412 |C|.. | . / |S-NEWS| \__ 413 | | . | . ( +------+ \ 414 +-+ . +---+ . ( +--------+ ) 415 ..| |.... \ |S-GAMING| ) 416 | P |..........( +--------+ ) 417 +-+ ...| | \ +---------+ ) 418 | | . +---+ ( |S-BANKING| / 419 |C|... | \_.+---------+ ) 420 | | | \.. / 421 +-+ / \____--' 422 / 423 Administrative / Internet 424 Domain, Zone 1 / 425 _________/ 427 Figure 1: Outbound TLS proxy 429 5.2.1. OP-5: Acceptable Use Policy (AUP) Enforcement (via payload 430 inspection) 432 Deployment: inline 434 A firewall or web proxy restricts a client in the same administrative 435 domain from accessing sites or services outside that domain per an 436 acceptable use policy. Similar in intent to OP-1, but the policy 437 enforcement in this practice requires access to data in the TLS 438 session (e.g., URL). 440 TLS 1.3 considerations: no capability impact. See Section 4.2 if a 441 selective decryption policy is used. 443 5.2.2. OP-6: Data Loss Prevention Compliance 445 Deployment: inline 447 A firewall enforces a Data Loss Prevention (DLP) policy by monitoring 448 the TLS sessions content of outbound communication for systems 449 sending organizational proprietary content or other restricted 450 information. Note that the firewall may be implemented and enforced 451 either at the endpoint or by the network infrastructure. 453 TLS 1.3 considerations: no capability impact. See Section 4.2 if a 454 selective decryption policy is used. 456 5.2.3. OP-7: Granular Network Segmentation 458 Deployment: inline 460 A firewall mediates the traffic between different policy zones in an 461 organization. The access policies between these zones may be based 462 on application names and categories rather than static IP addresses 463 and TCP/UDP port numbers. Through a TLS proxy, the firewall can 464 inspect URLs and other application parameters based on data in the 465 TLS session. 467 TLS 1.3 considerations: no capability impact. See Section 4.2 if a 468 selective decryption policy is used. 470 5.2.4. OP-8: Network-based Threat Protection (client) 472 Deployment: inline or out-of-band (depending on functionality) 474 Web proxies and firewalls protect end-users against a range of 475 threats by inspecting the data in the TLS session with a variety of 476 analytical techniques (e.g., signatures, heuristics, statistical 477 models, machine learning). This practice is a superset of OP-2. 478 Common goals are to prevent malware from reaching the endpoint, 479 preventing malware communication from a compromised host, restricting 480 lateral network movement of an intruder and gathering insight into 481 the behavior of threat activity on the network. 483 In certain deployments these technologies are also used to act as a 484 last line of defense against software vulnerabilities on endpoints - 485 either for 0-days for which there is no patch, or simply unpatched 486 clients. 488 TLS 1.3 considerations: no capability impact. See Section 4.2 if a 489 selective decryption policy is used. 491 5.2.5. OP-9: Protecting Challenging End Points 493 Deployment mode: inline 495 Web proxies, IPS and firewalls implement security policy and afford 496 protection to devices for which it is not feasible to run an end- 497 point solution (e.g., IoT); or that are end-of-life and will not 498 receive patches. This is a specialized instance of OP-8 targeting 499 these disadvantaged classes of devices. 501 These practices ensure that that older endpoints (and in some cases 502 even new ones) are not permanently vulnerable to newly discovered 503 vulnerabilities. 505 TLS 1.3 considerations: no capability impact. See Section 4.2 if a 506 selective decryption policy is used. 508 5.2.6. OP-10: Content Injection 510 Deployment: inline 512 A firewall or web proxy restricts message manipulation or insertion, 513 such as a block page or an interactive authentication portal 514 redirect, into the encrypted flow for the client to see. This may be 515 used in conjunction with OP-1, OP-5, and OP-7. 517 TLS 1.3 considerations: no capability impact. See Section 4.2 if a 518 selective decryption policy is used. 520 5.3. Inbound TLS Proxy 522 Inbound TLS proxy is the deployment scenario where the TLS proxy is 523 deployed in front of one or a set of servers or services. The 524 network device that implements the TLS proxy function is located in 525 the same administrative domain as the server(s) or service(s) it is 526 protecting. Usually it is not predictable or controllable as to 527 which TLS client will initiate a connection. See Figure 2. 529 The TLS proxy is provisioned with the server's certificates and 530 private keys so that it may either decrypt or terminate the TLS 531 connection on behalf of the server. In some instances, the TLS proxy 532 may periodically retrieve the private keys and associated 533 certificates from an external secure distribution service, such as a 534 HSM. Traffic between the TLS proxy and server may be encrypted or in 535 the clear; the former configuration is typical of a perimeter 536 firewall while the latter of a load-balancer. 538 ____________ 539 / 540 / S 541 _----__ / .--. 542 / \____ / |==| 543 __/ / |--| 544 / +-+ +-+ \__ | .....|==| S 545 ( | | | | \ | . |--| .--. 546 ( |C| +-+ |C| +-+ ) +---+ . |::| |==| 547 \ | | | | | | | | ) | |... |__| |--| 548 ( +-+ |C| +-+ |C|..............| P | S " " |==| 549 \ | | | | ) | |... .--. |--| 550 ( +-+ +-+ / +---+ . |==| |::| 551 \_. ) | . |--| |__| 552 \.. / | ..|==| " " 553 \____--' \ |--| 554 \ |::| Administrative 555 External Network \ |__| Domain 556 \ " " 557 \____________ 559 Figure 2: Inbound TLS proxy 561 5.3.1. OP-11: TLS offloading 563 Deployment mode: inline 565 Offloads crypto operations from the application server to a TLS 566 Proxy. This is not a typical security function on its own, but it 567 facilitates security control insertion downstream. As this is in the 568 same administrative domain, it is presumed that a TLS Proxy can be 569 provisioned with the appropriate keys when the TLS Server is 570 configured or managed. 572 TLS 1.3 considerations: no impact. 574 5.3.2. OP-12. Content distribution and application load balancing 576 Deployment mode: inline 578 Load balancers deployed in front of services provide resiliency 579 against denial of service attacks. TLS proxy functionality provides 580 access to the cleartext application layer data to enable service- 581 tailored load balancing. Similar to OP-11, it is presumed that a TLS 582 Proxy can be provisioned with the appropriate keys when the TLS 583 Server is configured or managed. 585 This practice may be combined with OP-11. 587 TLS 1.3 considerations: no impact. 589 5.3.3. OP-13: Network-based Threat Protection (server) 591 Deployment mode: inline and out-of-band 593 Web application firewalls (WAF) and firewalls protect servers and 594 services against a range of threats by inspecting the data in the TLS 595 session with a variety of analytical techniques (e.g., signatures, 596 heuristics, statistical models, machine learning). This practice is 597 identical in function to OP-8, but focused on threat prevention of 598 inbound requests to servers and services. 600 TLS 1.3 considerations for inline deployment mode: no capability 601 impact. Per Section 4.1, the network security device must explicitly 602 terminate the TLS connection from the client. 604 TLS 1.3 considerations for out-of-band mode: alternative approach 605 required. Per Section 4.1, active participation in the TLS exchange 606 is required to inspect the session. 608 5.3.4. OP-14: Full Packet Capture 610 Deployment mode: inline and out-of-band 612 A network security device stores a copy of all decrypted traffic that 613 meets a given filter. This traffic may be continuously captured in a 614 rolling buffer for use in future forensic analysis, incident 615 response, or computationally intensive retrospective analysis. This 616 collection may also be selectively enabled to support application 617 troubleshooting. 619 TLS 1.3 considerations for inline deployment mode: no capability 620 impact. Per Section 4.1, the network security device must explicitly 621 terminate the TLS connection from the client. 623 TLS 1.3 considerations for out-of-band mode: alternative approach 624 required. Per Section 4.1, offline decryption is not possible. 626 5.3.5. OP-15: Application Layer Gateway (ALG) 628 Deployment mode: inline 630 To conduct protocol conformance checks and rewrite embedded IP 631 addresses and TCP/UDP ports within the application layer payload for 632 traffic traversing a NAT boundary. While not strictly a security 633 function, this capability may typically be found in firewalls along 634 with the NAT supporting functions. 636 TLS 1.3 considerations: no impact. 638 6. Security Considerations 640 This document presents common and existing security monitoring and 641 detection functionality and how it interacts with TLS. It further 642 notes where existing practices will have to be adjusted to remain 643 effective as these solutions transition to include TLS 1.3 644 improvements. 646 These operational practices involve both good faith and malicious 647 client applications. The former category typically exhibits 648 consistently identifiable behavior and does not actively prevent any 649 transit inspection devices from performing application identification 650 for visibility and control purposes. The latter category of 651 applications actively attempts to circumvent network security 652 controls by deliberately manipulating various protocol headers, 653 injecting specific messages, and varying payload sizes in order to 654 avoid identification or to masquerade as a different permitted 655 application. 657 7. IANA Considerations 659 This document has no IANA actions. 661 8. Appendix A: Summary Impact to Operational Practices with TLS 1.3 662 +---------------------------------------------+-----------------------+ 663 | Operational Practice | Impact with TLS 1.3 | 664 +---------------------------------------------+-----------------------+ 665 | OP-1: AUP enforcement (headers only) | reduced effectiveness | 666 | OP-2: Behavior analytics (headers only) | reduced effectiveness | 667 | OP-3: Crypto compliance monitoring (server) | reduced effectiveness | 668 | OP-4: Crypto compliance monitoring (client) | reduced effectiveness | 669 | OP-5: AUP enforcement (payload) | no capability impact | 670 | OP-6: Data loss prevention compliance | no capability impact | 671 | OP-7: Granular network segmentation | no capability impact | 672 | OP-8: Network protection (client) | no capability impact | 673 | OP-9: Protecting challenging end points | no capability impact | 674 | OP-10: Content Injection | no capability impact | 675 | OP-11: TLS offloading | no impact | 676 | OP-12: Application load balancing | no impact | 677 | OP-13: inline: Network protection (server) | no operational impact | 678 | OP-13: oob: Network protection (server) | alternative required | 679 | OP-14: inline: Full packet capture | no operational impact | 680 | OP-14: oob: Full packet capture | alternative required | 681 | OP-15: Application layer gateway | no impact | 682 +---------------------------------------------+-----------------------+ 684 9. References 686 9.1. Normative References 688 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 689 Requirement Levels", BCP 14, RFC 2119, 690 DOI 10.17487/RFC2119, March 1997, 691 . 693 [RFC5246] Dierks, T. and E. Rescorla, "The Transport Layer Security 694 (TLS) Protocol Version 1.2", RFC 5246, 695 DOI 10.17487/RFC5246, August 2008, 696 . 698 [RFC7525] Sheffer, Y., Holz, R., and P. Saint-Andre, 699 "Recommendations for Secure Use of Transport Layer 700 Security (TLS) and Datagram Transport Layer Security 701 (DTLS)", BCP 195, RFC 7525, DOI 10.17487/RFC7525, May 702 2015, . 704 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 705 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 706 May 2017, . 708 [RFC8404] Moriarty, K., Ed. and A. Morton, Ed., "Effects of 709 Pervasive Encryption on Operators", RFC 8404, 710 DOI 10.17487/RFC8404, July 2018, 711 . 713 [RFC8446] Rescorla, E., "The Transport Layer Security (TLS) Protocol 714 Version 1.3", RFC 8446, DOI 10.17487/RFC8446, August 2018, 715 . 717 9.2. Informative References 719 [ECH] Rescorla, E., Oku, K., Sullivan, N., and C. Wood, "TLS 720 Encrypted Client Hello", draft-ietf-tls-esni-07 (work in 721 progress), June 2020. 723 [NONCE_FAIL] 724 Jovanovic, P., "Nonce-disrespecting adversaries: Practical 725 forgery attacks on GCM in TLS", 2016, 726 . 729 [TLS_VULNERABILITY] 730 Shenefiel, C., "PRNG Failures and TLS Vulnerabilities in 731 the Wild", 2017, 732 . 734 [WEAK_K2] Heninger, N., "Weak Keys Remain Widespread in Network 735 Devices", 2016, . 738 [WEAK_KEY] 739 Halderman, A., "Mining your Ps and Qs: Detection of 740 widespread weak keys in network devices", 2012, 741 . 744 Acknowledgments 746 The authors thank Andrew Ossipov, Flemming Andreasen, Kirsty Paine, 747 David McGrew, and Eric Vyncke for their contributions and valuable 748 feedback. 750 Authors' Addresses 752 Nancy Cam-Winget 753 Cisco Systems, Inc. 754 3550 Cisco Way 755 San Jose, CA 95134 756 USA 758 EMail: ncamwing@cisco.com 760 Eric Wang 761 Cisco Systems, Inc. 762 3550 Cisco Way 763 San Jose, CA 95134 764 USA 766 EMail: ejwang@cisco.com 768 Roman Danyliw 769 Software Engineering Institute 771 EMail: rdd@cert.org 773 Roelof DuToit 774 Broadcom 776 EMail: roelof.dutoit@broadcom.com