Lineage 2 Proxy: A Two-Leg Session, the Scope Decision and Closed Ports
The Lineage 2 client does not connect to a single address; it connects first to a login server and then to the world server you choose. This distinction drives the proxy decision: which leg gets routed depends on where you write the setting. This page explains scope, port blocks and the stability side in turn.
Two separate legsHow login server and world server connections differ under a proxy.
02
Scope decisionWhat router, operating system and application level each cover.
03
Port blocksPorts closed on campus and office networks, and the in-house route to a solution.
04
Latency budgetWhere latency accumulates and how to measure it.
The shortcut to understanding the relationship between Lineage 2 and a proxy is to see that the client talks to two separate servers. The login server authenticates the account and produces a short-lived session key; with that key the client opens a second connection to the world server you chose. If your rule covers only one of them, you get past the login screen but cannot enter the world.
The second point is the transport form: the game protocol establishes a single, long-lived session over TCP. That is good news for tunnelling, because both the CONNECT tunnel and SOCKS5 carry TCP. The same property lowers fault tolerance: a request that breaks in a browser is silently retried, whereas a connection that breaks here drops you back to the login screen.
The third is expectation management. Adding a hop lengthens the path; a proxy therefore increases latency in most setups rather than reducing it. The gain is not on the speed side but on the routing side: a fixed and predictable exit, a single address defined on a corporate network, verification of how content looks from another country.
From login server to world server: how is the session established?
The first connection made when you open the client goes to the login server and is short. The job is clear: the account is authenticated, the world list is fetched, and the client is given a temporary key to present to the server you choose. This leg does not stay open for the rest of the game; it closes once its job is done.
The moment you select a world, the client opens a new TCP connection; in most setups the target is both a different address and a different port. The common default of community server software is port 2106 for login and 7777 for the world server; in publisher-operated regions these values may differ. What matters is not the number but the fact that the two legs are separate from each other.
The third stage is the game itself. As your character moves, small packets flow continuously in both directions; the bandwidth demand is low, the continuity demand is high. That is why what you feel in Lineage 2 is less about average latency than about the variation in latency and the frequency of disconnections.
The fourth stage is disconnection, and it winds the loop back to the start. When you add a proxy, one more link joins the chain: a timeout at the exit, the end of a quota or a concurrent connection ceiling produces the same drop (concurrent connection limit).
Note
If login succeeds but you get stuck after character selection, the diagnosis can be made quickly: your rule covers the first leg but not the second. Define the scope by destination host, not by destination port.
DIAGRAMThe four stages of a Lineage 2 session
You can scroll the diagram horizontally to inspect it
The session is not a single connection but two separate legs and a return point; your proxy rule needs to cover both.
Which traffic enters the tunnel and which stays outside?
The client has no proxy field of its own
Unlike a browser, a game client has no field called "proxy server". The routing decision is therefore made outside the game: in the operating system setting, in a local routing application, or on the network device. Windows' system proxy setting is not a solution either; only applications that read that setting take it into account, and most game clients open their sockets directly (Windows proxy settings).
Protocol choice
An HTTP proxy opens a TCP tunnel using the CONNECT method to reach a non-HTTPS target; however, many providers restrict this method to well-known ports such as 80 and 443, and you are refused when you request a game port. SOCKS5 does not interpret the protocol it carries and can request an arbitrary destination port, so it is more compatible with game traffic (the difference between the two protocols).
Where the domain name is resolved also changes the outcome. If the client resolves the address on its own network, the target domain name is visible to your local DNS server and a record close to you may be returned; socks5h in this behaviour the proxy performs the resolution (where DNS is resolved in SOCKS5).
Out-of-game voice chat is a separate item. These applications mostly use UDP, and the CONNECT tunnel does not carry UDP; SOCKS5's UDP ASSOCIATE method can carry it, but both provider and client support are required (SOCKS5 UDP support).
Publisher regions, rented machines and community projects
Three separate hosting forms live side by side in the Lineage 2 ecosystem, and their network behaviours differ. Publisher-operated regional servers sit in large data centres, your account is tied to the region where it was created, and characters are not transferred between regions. On this path the route usually runs over mature backbones and changes very little from hour to hour.
The second group is community-operated servers. These are set up with open-source server software and often run on a single rented virtual or physical machine. A single location means no redundancy and no load distribution: if the server is in Frankfurt, the latency of a player connecting from a distant continent is fixed by distance, and no intermediate hop will shorten it.
The third is the case where you host the server yourself, and the question is reversed. In this scenario what you need is not an outbound exit but a front end that accepts incoming connections; when the two concepts are confused, the setup is planned wrongly from the start (the difference between forward and reverse proxies).
Server type
Typical hosting
Network outcome
The proxy's role
Publisher region
Multi-region data centre
Mature route, low variability
Fixed exit and regional view verification
Community server
A single rented machine
Distance fixes the latency
Routing only; it does not change distance
Your own server
A VPS you rent
You manage incoming traffic
The need is in the opposite direction: front end / reverse
Warning
Community servers may conflict with the publisher's terms of service. This page describes only network behaviour; which server you connect to and the consequences for your account are your responsibility.
Where do exit types fall in this game's scenarios?
Game traffic calls for a different balance than scraping or social media work. Pool size and rotation are irrelevant here; what you want is a single address that stays unchanged for a long time. So the decision axis should not be "how many IPs" but "how stable and how easy to set up".
Datacenter exits give the cheapest answer to this need: a fixed address, high bandwidth, simple setup. An ISP proxy offers a static address hosted in a provider's autonomous system; in terms of classification it sits close to a home connection, in terms of stability close to a data centre (ASN and IP reputation).
Residential and mobile exits are generally more complex than necessary for game sessions: the speed of the line is outside your control, the quota is billed on transferred data, and the address can change at an unexpected moment. Consider these not for the in-game connection but when examining the regional view of store and announcement pages.
Choose a single fixed exit and do not change it during the session.
Keep the exit at a reasonable distance from both you and the server.
Clarify the authentication method up front: IP authorisation or username and password?
Repeat the measurement with a ping test both with the proxy on and with it off.
DIAGRAMWhere exit types sit on the setup and stability axes
You can scroll the diagram horizontally to inspect it
The option closest to the top right corner is both easy to set up and stable throughout the session. The positions are a qualitative placement, not a measurement.
Choose a fixed exit for Lineage 2 sessions
In a game connection it is the stability of the address, not the breadth of the pool, that matters; ISP and datacenter solutions answer this need.
Choose whichever you need from our residential proxies, datacenter proxies, IPv6 and ISP solutions. Every plan comes with unlimited options, 99.9% uptime, rotating proxies, sticky sessions and 24/7 support. Ideal for web scraping, ad verification, SEO monitoring and digital data collection.
ISP ProxyStatic Turkish IPs registered to an ISP
ISP-registered static Türkiye IPs; they combine datacenter speed with the reputation of a real carrier. Ideal for long sessions and low-ping use.
Should the setting go on the router, the device or the application?
Router level
A rule written on the network device covers every device in the home, including the console, the television and a guest's phone. The breadth of that scope is also its biggest disadvantage: traffic you did not want routed is routed too, a single fault affects the whole household, and isolating the problem becomes harder. Moreover, most consumer routers do not speak SOCKS5; this setup requires either alternative firmware or an additional component running on the device (proxy via the router).
Device and application level
The operating system setting is the middle ground, but it runs into the limit noted earlier: only applications that read that setting are affected. The most precise result comes from the application level; a locally running routing component carries only the TCP connections of the process you select to the exit, while everything else takes its normal path.
Ask this before deciding: is there other work on this computer that you do not want routed? If the answer is yes, keep the scope narrow. Router level makes sense if you have a separate network segment whose sole purpose is gaming; not confusing address translation with the routing decision on the same device also matters at this point.
Where it is defined
Scope
Strength
Its difficulty
Router
Every device on the network
Also covers consoles and closed devices
Firmware support, broad side effects
Operating system
Applications that read the setting
The fastest path to set up
Game clients mostly ignore it
Application
The selected process
The narrowest and most predictable scope
A separate rule is needed for each process
Ports closed on campus and office networks
Most corporate and campus networks filter outbound traffic. The common policy is to allow only ports 80 and 443 outbound and close the rest. Because the game protocol does not listen on a well-known web port, it silently times out against this policy.
The symptom is familiar and easy to diagnose: the launcher downloads updates without trouble, because the download goes over HTTPS; but the login step hangs without giving any error. Two different kinds of traffic giving two different results tells you the block is at the network layer and port-based (what port numbers tell you).
The correct route to a solution runs through the institution. Tell the network administrator in writing which target you want to reach over which port; many institutions can open a defined exception for a specific source machine. If your provider offers an exit listening on an allowed port, that is another option that can be evaluated with the institution's approval. Using your own line outside working hours is often the most practical answer (access blocks on school and workplace networks).
Caution
Do not try to circumvent the network policy of the institution you belong to. On a corporate network the institution is the party that sets the rules; raising your access need with the administrator is both the correct and the lasting route.
Where does the latency budget accumulate?
The duration of a round is not produced in a single place. It starts in your own access network, continues on the leg to the exit, grows on the leg from the exit to the server, and comes back once the server has processed it. When you use a proxy you count three items instead of two; that is why the total almost always increases (what proxy latency is).
There is a rare exception, and it does not count as a rule: if your default route is needlessly circuitous, an exit that connects more directly to the server can shorten the total path. This should be measured, not assumed; a detailed discussion of the topic is in the article on whether a proxy lowers game ping.
In a game using a single long-lived connection, the handshake cost is paid only once; the real issue is how stable the path remains throughout the session. On a shared exit, the variation during peak hours can make the game noticeably stuttery without distorting the average value.
So do not trust a one-off measurement. Measure the same exit in the morning and in the evening, put the result side by side with the value obtained with the proxy off, and when deciding look not at the difference but at the consistency of the difference. Verify the exit's liveness separately with a proxy checker tool.
DIAGRAMWhere does the latency of a round accumulate?
You can scroll the diagram horizontally to inspect it
The proxy enlarges the total by adding the second and third items. The shares are relative weights; the percentages are not measured values.
From symptom to cause: common connection problems
Symptom
Possible cause
To be checked
Login goes through, the world does not
The rule covers only the first leg
Define the scope by destination host
The launcher updates, the game does not connect
The game port is closed on the network side
Clarify the network policy with the administrator
407 an authentication error of this kind
The username and password are not being sent
Verify the method and the IP authorisation
Disconnections at regular intervals
A timeout or session limit at the exit
Read the limits and remaining quota from the panel
The character responds slowly
The exit is shared and fluctuates at peak hours
Compare with a different exit at the same hour
The connection is never established
There is only an HTTP tunnel instead of SOCKS5
Ask which ports the provider allows
Authentication errors come from two sources: either the client is not sending the credentials at all, or the provider identifies you by IP authorisation and your home address has changed. The latter is common in the morning hours on lines with dynamic IPs.
Disconnections that recur at regular intervals are usually the result of policy rather than a fault. Some exits close an idle connection after a certain period; when the game sends no data for a long time, the tunnel silently drops and the client shows it as a connection error.
When is a proxy an unnecessary layer?
If you play from your own country, on your own line, with a single account, inserting a hop gains you nothing. The only measurable result you will get is a slightly longer path and one more suspect when a fault occurs.
If your aim is to push all of a device's traffic through a single tunnel, the tool you are looking for is different too. A proxy carries the scope you configure, not every connection on the system; this distinction is explained in detail in the article on the difference between a proxy and a VPN.
The meaningful use cases are narrow but clear: exiting a corporate network from a single defined address, using a second exit to determine whether a connection problem comes from your line or from the path in between, and verifying how announcement and store pages look in another country. For the equivalent of the same logic in other games, see the game proxy guides.
Frequently asked questions about Lineage 2 proxy
01Where do I enter a proxy address in the Lineage 2 client?
There is no such field in the client's own interface. Routing is defined outside the game: per process with a local routing application, through the operating system setting (which most games ignore), or on a network device. A per-process rule gives the most predictable result.
02I can log in but cannot enter the world — what is the cause?
Almost always a scope problem. After world selection, the client opens a new connection to a different address and port; if your rule covers only the login server, this second leg is not routed and times out.
03Should I prefer an HTTP proxy or SOCKS5 for gaming?
SOCKS5 is more suitable. An HTTP proxy tunnels to the target with CONNECT and providers often restrict this method to well-known web ports; you are refused when you request a game port. SOCKS5 does not interfere with the protocol it carries and can allow an arbitrary destination port.
04Does using a proxy improve in-game latency?
As a general rule, no; a hop is added and the path gets longer. In the rare cases where your default route is circuitous the total can shorten, but that is an exception and can only be determined by measurement. Base your decision on measurements taken with the proxy on and off, not on guesswork.
05Is it better to configure it on the router?
It depends on your scope requirement. A rule written on the network device covers every device including consoles, but it also carries traffic you did not want routed and makes fault isolation harder. If you do other things on the computer as well, a narrowly scoped rule is safer.
06What should I do if the game port is blocked on a school network?
Raising the need with the network administrator is the right path; many institutions can define an exception for a specific machine. Rather than trying to circumvent the institution's policy, using your own line outside working hours is the most practical solution in most cases.
07Are addresses from free lists suitable for this game?
Free lists are valuable for learning and testing, not for a game session. Most of these servers allow only web ports, who operates them is unknown, and they do not offer the stability required to carry a session that must stay open for hours.