Configuration reference / glossary

V2Ray Glossary

Put each client setting in the right layer: the protocol determines how to connect, routing determines where traffic goes, and DNS determines where domains are resolved.

Protocols & transports

When reviewing node settings, distinguish the proxy protocol, underlying transport, and transport security.

Compare protocols
Proxy protocols

VMess

VMess is a proxy protocol in the Project V ecosystem. It typically identifies users with a UUID and provides encryption at the protocol layer. It is configured separately from transports such as WebSocket. Even with matching UUIDs on both ends, a connection may fail if the transport or related settings don't match.

Proxy protocols

VLESS

VLESS is a proxy protocol often used with TLS or REALITY in node configurations. In a VLESS configuration, encryption: none means VLESS-layer encryption is disabled; it does not mean the entire connection is unencrypted. Check how the security layer is configured as well.

Proxy protocols

Trojan

Trojan is a password-authenticated proxy protocol that typically establishes connections over TLS. The password authorizes proxy access, while TLS verifies the server's identity. If you see a certificate name mismatch, check the server name and certificate instead of repeatedly re-entering the proxy password.

Transport security

REALITY

REALITY is a transport security mechanism provided by Xray. It is often used with VLESS and is not a separate proxy protocol. If an imported node appears as VLESS, that only means the protocol was recognized. A successful connection also requires a core that supports REALITY and reads the relevant handshake parameters.

Cores & clients

Clients store configuration; cores handle traffic forwarding. System proxy and TUN determine which traffic reaches the core.

View core configuration
Proxy core

Xray

Xray is a proxy core that handles protocols, routing, and outbound connections—not a client's graphical interface. The Android client v2rayNG uses Xray. Desktop v2rayN can use different cores depending on the node configuration, so seeing the same interface does not mean every node uses the same core.

Core projects

V2Fly

V2Fly is a community project that continues the technical direction of Project V. Its V2Ray core shares common roots with Xray. The Android client v2flyNG uses the V2Fly core. The cores support some of the same protocols, but advanced fields and transport settings are not necessarily interchangeable.

Traffic capture

System proxy

System proxy registers the client's local proxy address with the operating system for apps that follow system settings. A browser using the client does not mean command-line tools or other apps follow the same route. When troubleshooting an app, first check whether it uses system settings or its own proxy configuration.

Traffic capture

TUN mode

TUN mode receives IP packets through a virtual network interface, then lets the client take over and forward them according to its rules. It can cover apps that do not use the system proxy, but its actual scope depends on OS permissions, interface routes, and exclusions. If LAN devices become unreachable after enabling it, check whether the LAN subnet is being routed through the tunnel.

Routing

To determine where traffic goes, check the match criteria, rule order, and outbound tag in that order.

View routing rules
Match criteria

Routing rules

Routing rules match connections against conditions such as destination domain, IP, or port, then send matching traffic to a specified outbound. In common V2Ray and Xray configurations, rules are checked in order, and matching stops at the first hit. If a broad rule comes first, exceptions below it may never take effect.

Outbound selection

Traffic routing

Traffic routing sends different traffic to different outbounds—for example, direct connections for LAN traffic and proxy connections for selected domains. It determines where connections go; DNS routing, by contrast, selects a resolver. When troubleshooting, check which resolver handles the domain and which routing rule the connection ultimately matches.

Rule data

GeoIP

In routing rules, GeoIP refers to tagged collections of IP ranges. Their exact contents depend on the data files loaded by the client. GeoIP matches IP addresses, not domain names, and cannot pinpoint a server's physical location. If a connection is identified only by domain, whether it can be matched by IP also depends on the routing DNS strategy.

Rule data

GeoSite

GeoSite rules reference tagged collections of domain names. The same tag may contain different entries in different data files. Matching uses the destination domain known to the client, not the location associated with the server's IP. If the client sees only an IP and cannot recover the domain, a domain-set rule will not match.

DNS & resolution

Separate who resolves a domain, how the query is sent, and how the result is used for the connection.

View DNS settings
Resolution strategy

DNS routing

DNS routing chooses a resolver based on the queried domain—for example, sending an internal service domain to a server that can resolve it. The result is used for the subsequent connection, but choosing a resolver does not automatically change the connection's outbound. If an internal domain is unreachable, first check which address it resolved to, then check the connection route.

Query transport

DoH

DoH, or DNS over HTTPS, carries DNS queries and responses over HTTPS, protecting traffic between the requester and the resolver. The resolver can still see the domains it processes. Whether the connection to the resolver uses a proxy is determined separately by the client's DNS outbound settings or routing policy.

Address mapping

FakeDNS

FakeDNS returns a temporary mapped address for a domain and keeps the address-to-domain mapping inside the client. When a later connection follows the same traffic interception path, the core can recover the domain from the mapping and make a routing decision. If the mapped address is passed to a program without that mapping, the connection may fail.

Troubleshooting the DNS path

DNS leak

A DNS leak occurs when queries take a path outside the intended policy—for example, web traffic uses a proxy while domains are still resolved by the local network's resolver. Whether this counts as a leak depends on the configuration's intent: explicitly configured direct DNS resolution is not an unintended route. Also check whether the app has its own DNS settings enabled.

Subscriptions & nodes

Updating a subscription, checking node connectivity, and comparing request times are three separate checks.

View subscription management
Configuration sources

Subscription

A subscription is a source of node configurations that a client retrieves, usually by refreshing a subscription URL. A successful update means the client received and processed the subscription; it does not mean every node in it is reachable. Subscription URLs may contain access credentials, so avoid sharing the full URL in screenshots, logs, or exported files.

Connection configuration

Node

A node is a saved configuration for an outbound connection in a client, typically including the server address, port, protocol, and authentication parameters. Two entries can use the same server address but different transports or authentication settings. Compare the actual fields to determine whether they are equivalent; names in the list aren't enough.

Connection testing

Latency

Latency is the time a probe takes from start to result. What it measures—an ICMP response, a TCP connection, or an application request—depends on the client's test method. Results from different methods are not directly comparable. A test timeout only means that particular probe did not finish in time; it does not by itself prove the node is offline.

Connection testing

Real connection latency

Real connection latency usually measures the time taken for a request to a test address through the selected node. Compared with probing only the server port, it also covers the proxy handshake and outbound path, but results still depend on the test address and timeout settings. To compare nodes, use the same test address and method.

Security & encryption

When checking identities, distinguish the two sides: the proxy server authenticates the user, and the client verifies the server.

Transport protection

TLS

TLS provides encryption and integrity protection for a connection, and standard client configurations also verify the server's identity. It operates at a different layer from proxy protocols such as VMess and VLESS. Correct proxy credentials won't fix an expired certificate or a name mismatch. For these errors, check the certificate and system clock first.

Handshake parameters

SNI

SNI, or Server Name Indication, is sent during the TLS handshake to help a server at a shared address choose the right certificate for each domain. It is not the dial address or the HTTP Host field. Some clients set SNI using a server name field; if the connection address is an IP, that field may still need a domain name.

Client authentication

UUID

UUID is an identifier format used to identify and authenticate users in VMess or VLESS nodes. A correctly formatted UUID is not necessarily accepted by the server; the value entered in the client must match the one authorized by the server. Treat the full UUID as a credential when sharing screenshots or node details.

Server identity

TLS certificate validation

TLS certificate validation checks whether the peer's certificate is trusted, within its validity period, and matches the expected server name. When validation fails, first identify whether the issue is a name mismatch, an incorrect system time, or an untrusted certificate chain. Disabling validation as a routine setting removes an important layer of server identity protection.