Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

 
 

Next Gen Services Overview

This topic provides an overview of Next Gen Services and includes the following topics

5G Universal Router Services Overview

The 5G Universal routers support several types of Services interfaces, which provide specific capabilities for inspecting, monitoring and manipulating traffic as it transits a router. Services can be categorized into Adaptive Services and Next Gen Services, with each category providing Inline services interfaces and Multiservices interfaces options. Table 1 lists the cards that provide these services.

Note:

The MX-SPC3 replaces MS- type cards providing a significant overall performance improvement together with high-end scale and capacity.

Table 1: 5G Universal Router Services

5G Universal Routing Platform

Adaptive Services

Next Gen Services

MPC

si-1/0/0

Inline services

MS-DPC

sp-1/0/0

MS-MPC

ms-1/0/0

MS-MIC

ms-1/0/0

MPC

si-1/0/0

Inline services

MX-SPC3

vms-1/0/0

  • Adaptive Services can run on MS-DPC, MS-MPC, and MS-MIC cards using Multiservices (MS) PICs or Adaptive Services (AS) PICs.

  • Next Gen Services can run on MPC cards and the MX-SPC3 security services card.

Inline services are configured on Modular Port Concentrators (MPC)s. Inline services interfaces, are virtual physical interfaces that reside on the Packet Forwarding Engine. They provide high performance processing on traffic transiting the MPC, and allow you to maximize your chassis slot capacity and utilization.

Multiservices Security cards (MS-DPC, MS-MPC, MS-MIC or MX-SPC3), provide services that can be applied to any traffic transiting the chassis beyond just an individual MPC. They also provide dedicated processing to support a variety of security features at scale and high performance.

Adaptive Services Overview

Adaptive Services run inline on MPCs and on MS-DPC, MS-MPC, and MS-MIC Multiservice security cards. Adaptive Services (AS) PICs and Multiservices PICs enable you to perform multiple services on the same PIC by configuring a set of services and applications. The AS and Multiservices PICs offer a range of services that you can configure in one or more service sets.

Note:

On Juniper Networks 5G Universal Routing Platforms, the MS-DPC provides essentially the same capabilities as the MS-MPC. The interfaces on both platforms are configured in the same way.

For more information about Adaptive Services including inline services, see Adaptive Services Overview.

Inline Services

Adaptive Services also use inline services interfaces to provide inline services. Inline services interfaces are virtual interfaces that reside on the Packet Forwarding Engine.

You configure inline services only on MPCs using the naming convention si-fpc/pic/port rather than the ms-fpc/pic/port naming convention.

Next Gen Services

Next Gen Services provide the combined capabilities of security services enabling you to inspect, monitor and manipulate traffic as it transits the router. Next Gen Services are supported both inline on Modular Port Concentrators (MPCs) and the MX-SPC3 security services card in MX240, MX480 and MX960 routers that support this feature. Please refer to Table 2, which provides a summary of Next Gen Services that are supported both inline and on the MX-SPC3 card. Both Inline and MX-SPC3 based services can be used at the same time.

You configure Next Gen Services on the MX-SPC3 security services card using the virtual multiservices naming convention: vms-fpc/pic/port.

Summary of Services Supported on MX Series 5G Universal Routers

Table 2 provides a summary of the services supported under Next Gen Services.

Table 2: Summary of Services Supported on MX Series 5G Universal Routing Platform

Next Gen Services: Inline (si-) Interface and SPC3

Service Feature

Inline Services

SPC3

Junos OS Release

Sub-Service

Junos OS Release

Sub-Service

CGNAT

19.3R2

Basic-NAT44 and NAT66

Static Destination NAT

Twice-NAT44 Basic

6rd Softwires

NPTv6

19.3R2

Basic-NAT44

Basic-NAT66

Dynamic-NAT44

Static Destination NAT

Basic-NAT-PT

NAPT-PT

NAPT44

NAPT66

Port Block Allocation

Deterministic-nat44 and nat64

End Point Independent Mapping (EIM)/End Point Independent Filtering (EIF)

Persistent NAT – Application Pool Pairing (APP)

Twice-NAT44 – Basic, Dynamic and NAPT

NAT64

XLAT-464

NPTv6

20.1R1

Port Control Protocol (PCP) – v1 and v2

20.2R1

MAP-E

DS-Lite

NAT46

Traffic Load Balancer

19.3R2

 

19.3R2

 

SecIntel (ATP Cloud IP Threat Feeds)

19.3R2

 

N/A

 

Stateful Firewall Services

N/A

 

19.3R2

 

Intrusion Detection Services (IDS)

N/A

 

19.3R2

 

DNS Request Filtering

N/A

 

19.3R2

 

Aggregated Multiservices Interfaces

N/A

 

19.3R2

 

Inter-chassis High Availability

N/A

 

19.3R2

CGNAT, Stateful Firewall, IDS

URL Filtering

N/A

20.1R1

JFlow

20.1R1

N/A

RPM and TWAMP

20.1R1

N/A

Video Monitoring

20.1R1

N/A

IPsec VPN N/A   21.1R1

Route based Site 2 Site VPN

Traffic selector based VPNs

AutoVPN

Routing protocols (BGP/OSPF) over IPsec

Inline IPsec 24.2R1   N/A  

See Configuring IPsec VPN on MX-SPC3 Services Card for more information on IPsec support on SPC3 line card.

Next Gen Services Documentation

You can run Next Gen Services on the MX240, MX480, and MX960 routers that support this feature, if you have the SPC3 services card installed in the router. Refer to our TechLibrary for all router documentation. For Next Gen Services, refer to the following documentation:

Enabling Next Gen Services

To run Next Gen Services, you must enable it on the router. This enables the operating system to run it’s own operating system (OS) for Next Gen Services.

There are specific steps you’ll need to take if you’re migrating your services from legacy services cards to the SPC3. The Next Gen Services CLI differs from these legacy services. For more information, see Configuration Differences Between Adaptive Services and Next Gen Services.

Compatibility with Other Services Cards

The SPC3 services card is compatible end-to-end with the Switch Fabrics, Routing Engines and MS-MPC line cards as described in Table 3.

Table 3: SPC3 Services Card Compatibility with Switch Fabrics, Routing Engines and MPC Line Cards

Switch Fabric

Route Engine

MPC Line Cards

SCBE

RE-S-1800X4-16G-BB

RE-S-1800X4-16G-UPG-BB

RE-S-1800X4-16G-S

RE-S-1800X4-16G-R

RE-S-1800X4-32G-BB

RE-S-1800X4-32G-UB

RE-S-1800X4-32G-S

RE-S-1800X4-32G-R

MPC2E-3D

MPC2-3D-NG

MPC3E and MPC3E-3D-NG

MPC4E-3D

MPC-3D-16XGE

SCBE2

RE-S-1800X4-16G-BB

RE-S-1800X4-16G-UPG-BB

RE-S-1800X4-16G-S

RE-S-1800X4-16G-R

RE-S-1800X4-32G-BB

RE-S-1800X4-32G-UB

RE-S-1800X4-32G-S

RE-S-1800X4-32G-R

RE-S-X6-64G-BB

RE-S-X6-64G-UB

RE-S-X6-64G-S

RE-S-X6-64G-R

RE-S-X6-128G-S-BB

RE-S-X6-128G-S-S

RE-S-X6-128G-S-R

MPC2E-3D

MPC2-3D-NG

MPC3E and MPC3E-3D-NG

MPC4E-3D

MPC5E and MPC5EQ

MPC7E and MPC7EQ

MPC-3D-16XGE

SCBE3

RE-S-1800X4-16G-BB

RE-S-1800X4-16G-UPG-BB

RE-S-1800X4-16G-S

RE-S-1800X4-16G-R

RE-S-1800X4-32G-BB

RE-S-1800X4-32G-UB

RE-S-1800X4-32G-S

RE-S-1800X4-32G-R

RE-S-X6-64G-BB

RE-S-X6-64G-UB

RE-S-X6-64G-S

RE-S-X6-64G-R

RE-S-X6-128G-S-BB

RE-S-X6-128G-S-S

RE-S-X6-128G-S-R

MPC2-3D-NG

MPC3E-3D-NG

MPC4E-3D

MPC5E and MPC5EQ

MPC7E and MPC7EQ

MPC-3D-16XGE

MPC10E-10C

MPC10E-15C

Configuring the SPC3 Services Card

The interfaces on the SPC3 services card are referred to as a virtual multi service (vms) PIC. When you configure an SPC3 interface, you specify the interface as a vms- interface as follows:

Aside from the CLI differences, you need to be aware of the basic hardware differences between multiservices (MS) type (MS-DPC, MS-MPC, and MS-MIC) cards and the SPC3 services card. MS type cards contain four CPU complexes whereas the SPC3 card, while more powerful, contains two CPU complexes. Each CPU complex services a single PIC, meaning that MS type cards support four PICs whereas the SPC3 supports two PICs. MS type cards use special multiservices (MS) and adaptive services (AS) PICs, whereas the PICs on the SPC3 card are integrated.

Because the number of PICs directly affects the number of interfaces, you might need to add logical units to each interface on the SPC3 to increase the number of interfaces to four. For example, if you currently use all four interfaces on the MS type card and you have a service set per interface, you can create two logical units per interface on the SPC3 to bring the total number of interfaces to four, and then reassociate the four service sets to these four logical interfaces.

Methods for Applying Services to Traffic

When you configure Next Gen Services, you can apply those services with either of the following methods:

  • Apply the configured services to traffic that flows through a particular interface on the router.

  • Apply the configured services to traffic that is destined for a particular next hop.

Configuring IPsec VPN on SPC3 Services Card

To configuring IPsec on SPC3 service card, use the CLI configuration statements at the [edit security] hierarchy level as the IPsec CLI configuration at the [edit services] is replaced with the CLI configuration at the [edit security] hierarchy level as shown in Table 4

Table 4: Comparison on configuring IPsec VPN for MX and MX-SPC3
Current MX Configuration Equivalent MX-SPC3 Configuration
set services ipsec-vpn traceoptions set security ike traceoptions
set services ipsec-vpn ike proposal set security ike proposal
set services ipsec-vpn ike policy set security ike policy
set services ipsec-vpn ike policy policy-name respond-bad-spi set security ike respond-bad-spi
set services ipsec-vpn ipsec proposal set security ipsec proposal
set services ipsec-vpn ipsec policy set security ipsec policy
set services ipsec-vpn rule rule-name term term-name from [source-address| destination-address] set security ipsec vpn vpn-name traffic-selector selector-name [local-ip | remote-ip]
set services ipsec-vpn rule rule-name term term-name from ipsec-inside-interface set security ipsec vpn vpn-name bind-interface
set services ipsec-vpn rule rule-name term term-name then remote-gateway set security ike gateway gw-name address
set services ipsec-vpn rule rule-name term term-name then backup-remote-gateway set security ike gateway gw-name address
set services ipsec-vpn rule rule-name term term-name then dead-peer-detection set security ike gateway gw-name dead-peer-detection
set services ipsec-vpn rule rule-name term term-name then dynamic ike-policy set security ike gateway gw-nameike-policy
set services ipsec-vpn rule rule-name term term-name then dynamic ipsec-policy set security ipsec vpn vpn-name ike ipsec-policy
set services ipsec-vpn rule rule-name term term-name then manual set security ipsec vpn vpn-name manual
set services ipsec-vpn rule rule-name term term-name then clear-dont-fragment-bit set security ipsec vpn vpn-name df-bit clear
set services ipsec-vpn rule rule-name term term-name then copy-dont-fragment-bit set security ipsec vpn vpn-name df-bit copy
set services ipsec-vpn rule rule-name term term-name then set-dont-fragment-bit set security ipsec vpn vpn-name df-bit copy
set services ipsec-vpn rule rule-name term term-name then tunnel-mtu set security ipsec vpn vpn-name tunnel-mtu
set services ipsec-vpn rule rule-name term term-name then no-anti-replay set security ipsec vpn vpn-name ike no-anti-replay
set services ipsec-vpn rule rule-name match-direction set security ipsec vpn vpn-namematch-direction
set services ipsec-vpn establish-tunnels set security ipsec vpn vpn-nameestablish-tunnels
set services service-set svc-set-name ipsec-vpn-options local-gateway address set security ipsec vpn vpn-nameike gateway gateway-name
set services service-set svc-set-name ipsec-vpn-options clear-dont-fragment-bit No global service-set setting. Must be configured on a per vpn object basis.
set services service-set svc-set-name ipsec-vpn-options copy-dont-fragment-bit No global service-set setting. Must be configured on a per vpn object basis.
set services service-set svc-set-name ipsec-vpn-options set-dont-fragment-bit No global service-set setting. Must be configured on a per vpn object basis.
set services service-set svc-set-name ipsec-vpn-options udp-encapsulate set security ipsec vpn vpn-nameudp-encapsulate
set services service-set svc-set-name ipsec-vpn-options no-anti-replay No global service-set setting. Must be configured on a per vpn object basis.
set services service-set svc-set-name ipsec-vpn-options passive-mode-tunneling set security ipsec vpn vpn-name passive-mode-tunneling
set services service-set svc-set-name ipsec-vpn-options tunnel-mtu No global service-set setting. Must be configured on a per vpn object basis.
set services service-set svc-set-name ipsec-vpn-rules set services service-set svc-set-name ipsec-vpn-rules
set services ipsec-vpn rule <rule-name> term <term-name> then tunnel-mtu

set security ipsec vpn <vpn-name> tunnel-mtu

Understanding Tunnel MTU

Tunnel MTU (Maximum Transmission Unit) is the maximum packet size that can be sent through a tunnel without fragmentation. For a route-based IPsec tunnel there are two independent fragmentation points on the clear-text packet and one for the encrypted packet:

  • st0 (secure tunnel) MTU is enforced by the PFE on clear-text packets before they are forwarded to the service card. When a packet exceeds the configured st0 MTU, the PFE fragments it before forwarding. The packet is divided into approximately equal-sized fragments, creating as many fragments as necessary, rather than maximizing the size of each fragment up to the MTU limit.
  • tunnel-MTU is enforced by flowd (flow daemon), the packet-processing component responsible for stateful flow handling and IPsec services, during packet encryption. Before encrypting a clear-text packet or fragment received from the PFE, flowd verifies that its size, including IPsec overhead, does not exceed the configured tunnel-MTU. If the packet exceeds the limit, flowd fragments it into smaller packets that fit within the tunnel-MTU and then encrypts each fragment.

  • External (physical) interface MTU is applied to the encrypted packet after IPsec encapsulation on egress.

Note:
  • The configured tunnel-MTU must be less than or equal to the st0 MTU; otherwise, the configuration commit fails.

  • When the PFE performs pre-fragmentation, flowd processes the resulting fragments only for session lookup and validation. It does not reassemble the fragments into the original packet. Instead, each fragment is processed independently by flowd.

Difference between st0 MTU and tunnel MTU

  • Tunnel-MTU is at different level compared to st0 MTU.
  • st0 MTU is interface level MTU and tunnel-MTU feature achieves tunnel level MTU
  • In MX-SPC3, PFE checks st0 mtu to fragment or drop the packet. Hence, packet does not reach flowd or IPsec and will not have any control over the MTU action.
  • VPN tunnel-mtu configuration value is less than the st0 MTU.

Examples

For the examples that follow, assume a clear-text packet size of about 1862 bytes and IPsec overhead of 60–100 bytes. Therefore, the effective clear-text payload supported by a configured MTU of M is roughly M − overhead.

Case 1: Secure Tunnel MTU > Packet Size (Only tunnel-MTU Causes Fragmentation)

Assume st0 MTU = 9192 (default), tunnel-MTU = 1400, packet size ≈ 1862 bytes. As the packet size is smaller than the st0 MTU, the PFE does not fragment the packet. The packet reaches flowd intact. flowd compares the packet size against the effective tunnel-MTU limit (approximately 1300 bytes after accounting for IPsec overhead). Since the packet exceeds the tunnel-MTU limit, flowd fragments it into appropriately sized pieces, for example, 1300 bytes and 562 bytes. flowd encrypts each fragment and forwards it for transmission.

As a result only one fragmentation operation occurs, and it is performed by flowd based on the tunnel-MTU setting. This is the preferred behavior because tunnel-MTU serves as the sole fragmentation control point, allowing fragments to be sized efficiently to maximize payload utilization.

Case 2: Secure Tunnel MTU < Packet Size and PFE Fragments Still Exceed tunnel-MTU (Double Fragmentation)

Assume st0 MTU = 1500, tunnel-MTU = 700, packet size ≈ 1862 bytes. As the packet exceeds the st0 MTU, the PFE performs the first fragmentation step, splitting the packet into approximately equal-sized fragments (for example, 936 bytes and 926 bytes). These fragments are then passed to flowd for processing. flowd evaluates each fragment against the effective tunnel-MTU limit (approximately 600 bytes after accounting for IPsec overhead). Since both PFE-generated fragments still exceed the tunnel-MTU limit, flowd fragments each one again:

936 → 600 + 336

926 → 600 + 326

flowd processes each fragment independently. It does not reassemble the PFE-generated fragments before applying tunnel-MTU fragmentation.

As a result two fragmentation stages occur- first by the PFE based on the st0 MTU, and then by flowd based on the tunnel-MTU. The original packet is ultimately transmitted as four encrypted fragments. This is an undesirable scenario as it introduces additional fragmentation overhead and reduces efficiency.

Case 3: Secure Tunnel MTU < Packet Size and PFE Fragments Are Smaller Than tunnel-MTU (No Additional tunnel-MTU Fragmentation)

Assume st0 MTU = 1500, tunnel-MTU = 1438, packet size ≈ 1862 bytes. As the packet size exceeds the st0 MTU, the PFE fragments the packet into approximately equal-sized fragments, for example, 936 bytes and 926 bytes. These fragments are then forwarded to flowd for IPsec processing. flowd compares each fragment against the effective tunnel-MTU limit (approximately 1378 bytes after accounting for IPsec overhead). Since both fragments are already smaller than the tunnel-MTU limit, flowd does not perform any additional fragmentation. flowd encrypts and forwards the fragments unchanged.

As a result only one fragmentation step occurs, and it is performed by the PFE based on the st0 MTU. Although a tunnel-MTU is configured, it has no effect because the PFE-generated fragments already fit within the tunnel-MTU limit. In this scenario, the st0 MTU, rather than the tunnel-MTU, determines the resulting fragment sizes.

Note: Set the st0 MTU high enough to prevent PFE fragmentation and use tunnel-MTU as the primary fragmentation control. This enables a single flowd-driven fragmentation step and avoids inefficient double fragmentation.

Significace of Each MTU Setting

  • External interface MTU determines the maximum size of the encrypted packet that can be transmitted on the physical network. It is enforced on egress after IPsec encapsulation.

  • st0 MTU is an interface-level MTU shared by all VPNs and tunnels associated with the st0 interface. It is enforced by the PFE on clear-text traffic before encryption. If a packet exceeds the st0 MTU, the PFE fragments it before forwarding it to flowd.

  • tunnel-MTU is a per-VPN setting enforced by flowd. It provides a VPN-specific packet size limit without affecting other VPNs that share the same st0 interface. tunnel-MTU only applies to packets or fragments that still exceed its limit when they reach flowd.

  • tunnel-MTU must be less than or equal to the st0 MTU. When both settings are configured, st0 MTU controls PFE fragmentation, while tunnel-MTU controls VPN-specific fragmentation performed by flowd.