Quick reference

V2Ray Glossary

Look up terms you see in the client interface, connection settings, or logs. Identify which layer a term belongs to before deciding what to check.

Protocol
How the two ends of a connection exchange data
Traffic capture
How traffic enters the client
Routing
Which outbound route traffic takes after entering the core

Protocols & Transport

Protocols, transport methods, and security layers are separate settings. Matching the protocol name alone does not mean the other parameters are compatible.

VMess

VMess is an early proxy protocol used by Project V. It authenticates with a user ID and defines data protection at the protocol level. Transport methods such as WebSocket are configured separately, so the name VMess does not tell you which transport is in use. If the settings look correct but the handshake fails, check whether the device clock is synchronized.

VLESS

VLESS and VMess are different proxy protocols, and VLESS configurations commonly use a UUID to identify the user. Deployments often pair it with TLS or REALITY as the connection security layer. Changing the protocol name does not convert an existing VMess node to VLESS; both ends must support and use compatible settings.

Trojan

Trojan is a proxy protocol commonly paired with TLS and typically uses a password for proxy-level authentication. TLS certificate verification checks the server's identity, while the proxy password authenticates the user; neither replaces the other. If the logs report a certificate error first, check the server name and certificate status before changing the password.

REALITY

REALITY is a connection security mechanism supported by Xray and commonly paired with VLESS. When the interface lists VLESS and REALITY separately, the former is the proxy protocol and the latter is the connection security setting. If a node imports but will not start, check whether the selected core and version support those fields.

Core & Client

The interface manages settings; the core handles connections. Recording the version of each is more useful than noting only the client name.

V2Fly

V2Fly is an open-source community project that continues development of the V2Ray core from Project V. Client interfaces may also use the name for the core maintained by the project. It is not a type of share link. If documentation says a field applies to Xray, do not assume a V2Fly field with the same name behaves the same way.

Xray

Xray is an independent proxy core that grew out of the V2Ray ecosystem and supports features such as REALITY. It shares some configuration concepts with V2Fly, but protocol capabilities and supported fields vary by version. If you see an unknown-field error, check the startup logs to confirm which core is actually loaded.

Client

v2rayN, v2rayNG, and v2flyNG are graphical clients for managing subscriptions, selecting nodes, switching modes, and viewing logs. A client may first save imported content as a list, then generate a configuration for the active connection. An entry appearing in the list only confirms that import produced a result; it does not prove the remote connection is reachable.

Core

The core reads the configuration generated by the client and handles protocol handshakes, connection forwarding, and route matching. If the interface opens but the core fails to start, check the first error in the logs; node latency is less relevant. If the logs say a local port cannot be bound, switching remote nodes usually will not resolve the port conflict.

Subscriptions & Nodes

A subscription update, an accessible port, and a successful proxy request are three separate things. When troubleshooting, first note which step failed.

Subscription

A subscription lets a client retrieve multiple node entries from one URL. It can be refreshed manually or, in some clients, on a schedule. A successful update means the list was retrieved and parsed, not that every node is reachable. If an update fails, determine whether the request timed out or returned a response that contained no parseable entries.

Node

A node is a remote connection profile in the client, typically containing an address, port, protocol, authentication details, and transport settings. The name in the list is just a label; entries with the same name may point to different addresses or ports. Before comparing connection results, confirm that you selected the same profile, not just one with the same name.

TCPing

TCPing attempts to establish a TCP connection to a specified address and port, and records how long it takes. A successful test confirms that a TCP connection could be established at that moment; it does not verify proxy credentials or prove that a destination website will load. This test alone cannot determine whether a connection using UDP is working properly.

Real connection latency

Real connection latency measures a request sent through a specified node to a test destination. Results depend on the protocol handshake, test destination, and timing method. It checks more of the connection than a port-only test, but it is not a download speed test and does not mean every website is reachable. For node comparisons, use the same test destination and network conditions where possible.

Routing & Split Routing

First confirm that traffic reaches the core, then check which rule it matches. Changing rules cannot reroute connections that never entered the core.

Routing rule

Routing rules map matching conditions, such as a domain, IP address, or port, to a specified outbound. Common cores check rules in configuration order, with the first match taking effect. If a new exception seems to have no effect, check whether it comes after a broader rule that already matches the traffic.

Split routing

Split routing sends different connections to different outbounds, such as direct, proxy, or blocked. A webpage and its images may come from different domains, so they can match different rules. If the page loads but images do not, note the destination of the failed requests instead of checking only the domain shown in the address bar.

GeoIP

GeoIP is a collection of IP address ranges that routing rules can reference. For example, geoip:private is commonly used to match private address ranges. It matches addresses; it does not resolve domains to IPs. Whether a domain connection matches this kind of rule also depends on whether the core has the destination IP available when it evaluates the rule.

GeoSite

GeoSite organizes domains into reusable rule collections by label, commonly referenced as geosite: followed by a label. Which labels exist and which domains they include depend on the rule resource actually loaded. If an app provides only the destination IP and the core has not obtained the domain, these rules cannot match the IP directly.

Networking & DNS

Web requests and domain lookups may take different routes. Even after confirming the proxy exit, check DNS routing separately.

System proxy

System proxy is an operating-system setting that apps can read; graphical clients often point it to a local listening address. An app with its own proxy setting may override the system setting or continue connecting directly. If one browser works but another app does not, first check whether the latter uses the system proxy.

TUN mode

TUN mode uses a virtual network interface to receive IP traffic covered by system routes, without requiring every app to read the system proxy setting. The routing table, exclusions, and platform permissions determine which traffic actually enters the interface. If local network services become unreachable after enabling TUN, check whether the local subnet is included in the captured traffic.

FakeDNS

FakeDNS returns a mapped address to an app. When a later connection reaches the core, the core can use that address to recover the original domain and route the connection according to the rules. The mapped address is not the destination server's real IP. After changing traffic capture or DNS settings, an app with an old address cached may be unable to resolve the original mapping for new connections.

DNS leak

A DNS leak occurs when a query that should follow a specified resolution policy takes an unexpected route or is sent to another resolver. A website using the expected proxy exit does not, by itself, prove that DNS queries followed the correct path. When diagnosing the issue, identify whether the query came from the browser, the operating system, or the core, then compare its route with the intended DNS path.

Troubleshoot by symptom

Back to category index ↑