HttpClient has a complete control over the process of connection initialization and termination as well as I/O operations on active connections. However various aspects of connection operations can be controlled using a number of parameters.

These are parameters that can influence connection operations: defines the socket timeout (SO_TIMEOUT) in milliseconds, which is the timeout for waiting for data or, put differently, a maximum period inactivity between two consecutive data packets). A timeout value of zero is interpreted as an infinite timeout. This parameter expects a value of type java.lang.Integer. If this parameter is not set read operations will not time out (infinite timeout). determines whether Nagle's algorithm is to be used. The Nagle's algorithm tries to conserve bandwidth by minimizing the number of segments that are sent. When applications wish to decrease network latency and increase performance, they can disable Nagle's algorithm (that is enable TCP_NODELAY. Data will be sent earlier, at the cost of an increase in bandwidth consumption. This parameter expects a value of type java.lang.Boolean. If this parameter is not, TCP_NODELAY will be enabled (no delay). determines the size of the internal socket buffer used to buffer data while receiving / transmitting HTTP messages. This parameter expects a value of type java.lang.Integer. If this parameter is not set HttpClient will allocate 8192 byte socket buffers. sets SO_LINGER with the specified linger time in seconds. The maximum timeout value is platform specific. Value 0 implies that the option is disabled. Value -1 implies that the JRE default is used. The setting only affects the socket close operation. If this parameter is not set value -1 (JRE default) will be assumed. determines the timeout in milliseconds until a connection is established. A timeout value of zero is interpreted as an infinite timeout. This parameter expects a value of type java.lang.Integer. If this parameter is not set connect operations will not time out (infinite timeout). determines whether stale connection check is to be used. Disabling stale connection check may result in a noticeable performance improvement (the check can cause up to 30 millisecond overhead per request) at the risk of getting an I/O error when executing a request over a connection that has been closed at the server side. This parameter expects a value of type java.lang.Boolean. For performance critical operations the check should be disabled. If this parameter is not set the stale connection will be performed before each request execution. determines the maximum line length limit. If set to a positive value, any HTTP line exceeding this limit will cause an java.io.IOException. A negative or zero value will effectively disable the check. This parameter expects a value of type java.lang.Integer. If this parameter is not set, no limit will be enforced. determines the maximum HTTP header count allowed. If set to a positive value, the number of HTTP headers received from the data stream exceeding this limit will cause an java.io.IOException. A negative or zero value will effectively disable the check. This parameter expects a value of type java.lang.Integer. If this parameter is not set, no limit will be enforced. defines the maximum number of ignorable lines before we expect a HTTP response's status line. With HTTP/1.1 persistent connections, the problem arises that broken scripts could return a wrong Content-Length (there are more bytes sent than specified). Unfortunately, in some cases, this cannot be detected after the bad response, but only before the next one. So HttpClient must be able to skip those surplus lines this way. This parameter expects a value of type java.lang.Integer. 0 disallows all garbage/empty lines before the status line. Use java.lang.Integer#MAX_VALUE for unlimited number. If this parameter is not set unlimited number will be assumed.

The process of establishing a connection from one host to another is quite complex and involves multiple packet exchanges between two endpoints, which can be quite time consuming. The overhead of connection handshaking can be significant, especially for small HTTP messages. One can achieve a much higher data throughput if open connections can be re-used to execute multiple requests. HTTP/1.1 states that HTTP connections can be re-used for multiple requests per default. HTTP/1.0 compliant endpoints can also use similar mechanism to explicitly communicate their preference to keep connection alive and use it for multiple requests. HTTP agents can also keep idle connections alive for a certain period time in case a connection to the same target host may be needed for subsequent requests. The ability to keep connections alive is usually refered to as connection persistence. HttpClient fully supports connection persistence.

HttpClient is capable of establishing connections to the target host either directly or via a route that may involve multiple intermediate connections also referred to as hops. HttpClient differentiates connections of a route into plain, tunneled and layered. The use of multiple intermediate proxies to tunnel connections to the target host is referred to as proxy chaining. Plain routes are established by connecting to the target or the first and only proxy. Tunnelled routes are established by connecting to the first and tunnelling through a chain of proxies to the target. Routes without a proxy cannot be tunnelled. Layered routes are established by layering a protocol over an existing connection. Protocols can only be layered over a tunnel to the target, or over a direct connection without proxies.

RouteInfo interface represents information about a definitive route to a target host involving one or more intermediate steps or hops. HttpRoute is a concrete implementation of RouteInfo, which cannot be changed (is immutable). HttpTracker is a mutable RouteInfo implementation used internally by HttpClient to track the remaining hops to the ultimate route target. HttpTracker can be updated after a successful execution of the next hop towards the route target. HttpRouteDirector is a helper class that can be used to compute the next step in a route. This class is used internally by HttpClient. HttpRoutePlanner is an interface representing a strategy to compute a complete route to a given target based on the execution context. HttpClient ships with two default HttpRoutePlanner implementation. ProxySelectorRoutePlanner is based on java.net.ProxySelector. By default, it will pick up the proxy settings of the JVM, either from system properties or from the browser running the application. DefaultHttpRoutePlanner implementation does not make use of any Java system properties, nor of system or browser proxy settings. It computes routes based exclusively on HTTP parameters described below.

HTTP connections can be considered secure if information transmitted between two connection endpoints cannot be read or tampered with by an unauthorized third party. The SSL/TLS protocol is the most widely used technique to ensure HTTP transport security. However, other encryption techniques could be employed as well. Usually, HTTP transport is layered over the SSL/TLS encrypted connection.

These are parameters that can influence route computation: defines a proxy host to be used by default route planners that do not make use of JRE settings. This parameter expects a value of type HttpHost. If this parameter is not set direct connections to the target will be attempted. defines a local address to be used by all default route planner. On machines with multiple network interfaces, this parameter can be used to select the network interface from which the connection originates. This parameter expects a value of type java.net.InetAddress. If this parameter is not set a default local address will be used automatically. defines an forced route to be used by all default route planner. Instead of computing a route, the given forced route will be returned, even if it points to a completely different target host. This parameter expects a value of type HttpRoute.

HTTP connections make use of a java.net.Socket object internally to handle transmission of data across the wire. They, however, rely on SchemeSocketFactory interface to create, initialize and connect sockets. This enables the users of HttpClient to provide application specific socket initialization code at runtime. PlainSocketFactory is the default factory for creating and initializing plain (unencrypted) sockets. The process of creating a socket and that of connecting it to a host are decoupled, so that the socket could be closed while being blocked in the connect operation.

LayeredSchemeSocketFactory is an extension of SchemeSocketFactory interface. Layered socket factories are capable of creating sockets layered over an existing plain socket. Socket layering is used primarily for creating secure sockets through proxies. HttpClient ships with SSLSocketFactory that implements SSL/TLS layering. Please note HttpClient does not use any custom encryption functionality. It is fully reliant on standard Java Cryptography (JCE) and Secure Sockets (JSEE) extensions.

HttpClient makes use of SSLSocketFactory to create SSL connections. SSLSocketFactory allows for a high degree of customization. It can take an instance of javax.net.ssl.SSLContext as a parameter and use it to create custom configured SSL connections. Customization of SSLSocketFactory implies a certain degree of familiarity with the concepts of the SSL/TLS protocol, a detailed explanation of which is out of scope for this document. Please refer to the Java Secure Socket Extension for a detailed description of javax.net.ssl.SSLContext and related tools.

In addition to the trust verification and the client authentication performed on the SSL/TLS protocol level, HttpClient can optionally verify whether the target hostname matches the names stored inside the server's X.509 certificate, once the connection has been established. This verification can provide additional guarantees of authenticity of the server trust material. X509HostnameVerifier interface represents a strategy for hostname verification. HttpClient ships with three X509HostnameVerifier. Important: hostname verification should not be confused with SSL trust verification. The strict hostname verifier works the same way as Sun Java 1.4, Sun Java 5, Sun Java 6. It's also pretty close to IE6. This implementation appears to be compliant with RFC 2818 for dealing with wildcards. The hostname must match either the first CN, or any of the subject-alts. A wildcard can occur in the CN, and in any of the subject-alts. The hostname verifier that works the same way as Curl and Firefox. The hostname must match either the first CN, or any of the subject-alts. A wildcard can occur in the CN, and in any of the subject-alts. The only difference between BrowserCompatHostnameVerifier and StrictHostnameVerifier is that a wildcard (such as "*.foo.com") with BrowserCompatHostnameVerifier matches all subdomains, including "a.b.foo.com". This hostname verifier essentially turns hostname verification off. This implementation is a no-op, and never throws the javax.net.ssl.SSLException. Per default HttpClient uses BrowserCompatHostnameVerifier implementation. One can specify a different hostname verifier implementation if desired

Scheme class represents a protocol scheme such as "http" or "https" and contains a number of protocol properties such as the default port and the socket factory to be used to creating java.net.Socket instances for the given protocol. SchemeRegistry class is used to maintain a set of Schemes HttpClient can choose from when trying to establish a connection by a request URI:

Even though HttpClient is aware of complex routing scemes and proxy chaining, it supports only simple direct or one hop proxy connections out of the box. The simplest way to tell HttpClient to connect to the target host via a proxy is by setting the default proxy parameter: One can also instruct HttpClient to use standard JRE proxy selector to obtain proxy information: Alternatively, one can provide a custom RoutePlanner implementation in order to have a complete control over the process of HTTP route computation:

Operated connections are client side connections whose underlying socket or its state can be manipulated by an external entity, usually referred to as a connection operator. OperatedClientConnection interface extends HttpClientConnection interface and define additional methods to manage connection socket. The ClientConnectionOperator interface represents a strategy for creating OperatedClientConnection instances and updating the underlying socket of those objects. Implementations will most likely make use SchemeSocketFactorys to create java.net.Socket instances. The ClientConnectionOperator interface enables the users of HttpClient to provide a custom strategy for connection operators as well as an ability to provide alternative implementation of the OperatedClientConnection interface.

HTTP connections are complex, stateful, thread-unsafe objects which need to be properly managed to function correctly. HTTP connections can only be used by one execution thread at a time. HttpClient employs a special entity to manage access to HTTP connections called HTTP connection manager and represented by the ClientConnectionManager interface. The purpose of an HTTP connection manager is to serve as a factory for new HTTP connections, manage persistent connections and synchronize access to persistent connections making sure that only one thread can have access to a connection at a time. Internally HTTP connection managers work with instances of OperatedClientConnection, but they hands out instances of ManagedClientConnection to the service consumers. ManagedClientConnection acts as a wrapper for a OperatedClientConnection instance that manages its state and controls all I/O operations on that connection. It also abstracts away socket operations and provides convenience methods for opening and updating sockets in order to establish a route. ManagedClientConnection instances are aware of their link to the connection manager that spawned them and of the fact that they must be returned back to the manager when no longer in use. ManagedClientConnection classes also implement ConnectionReleaseTrigger interface that can be used to trigger the release of the connection back to the manager. Once the connection release has been triggered the wrapped connection gets detached from the ManagedClientConnection wrapper and the OperatedClientConnection instance is returned back to the manager. Even though the service consumer still holds a reference to the ManagedClientConnection instance, it is no longer able to execute any I/O operation or change the state of the OperatedClientConnection either intentionally or unintentionally. This is an example of acquiring a connection from a connection manager: The connection request can be terminated prematurely by calling ClientConnectionRequest#abortRequest() if necessary. This will unblock the thread blocked in the ClientConnectionRequest#getConnection() method. BasicManagedEntity wrapper class can be used to ensure automatic release of the underlying connection once the response content has been fully consumed. HttpClient uses this mechanism internally to achieve transparent connection release for all responses obtained from HttpClient#execute() methods:

SingleClientConnManager is a simple connection manager that maintains only one connection at a time. Even though this class is thread-safe it ought to be used by one execution thread only. SingleClientConnManager will make an effort to reuse the connection for subsequent requests with the same route. It will, however, close the existing connection and open it for the given route, if the route of the persistent connection does not match that of the connection request. If the connection has been already been allocated java.lang.IllegalStateException is thrown. SingleClientConnManager is used by HttpClient per default.

ThreadSafeClientConnManager is a more complex implementation that manages a pool of client connections and is able to service connection requests from multiple execution threads. Connections are pooled on a per route basis. A request for a route which already the manager has persistent connections for available in the pool will be services by leasing a connection from the pool rather than creating a brand new connection. ThreadSafeClientConnManager maintains a maximum limit of connection on a per route basis and in total. Per default this implementation will create no more than 2 concurrent connections per given route and no more 20 connections in total. For many real-world applications these limits may prove too constraining, especially if they use HTTP as a transport protocol for their services. Connection limits, however, can be adjusted using HTTP parameters. This example shows how the connection pool parameters can be adjusted:

When an HttpClient instance is no longer needed and is about to go out of scope it is important to shut down its connection manager to ensure that all connections kept alive by the manager get closed and system resources allocated by those connections are released.

When equipped with a pooling connection manager such as ThreadSafeClientConnManager HttpClient can be used to execute multiple requests simultaneously using multiple threads of execution. ThreadSafeClientConnManager will allocate connections based on its configuration. If all connections for a given route have already been leased, a request for connection will block until a connection is released back to the pool. One can ensure the connection manager does not block indefinitely in the connection request operation by setting 'http.conn-manager.timeout' to a positive value. If the connection request cannot be serviced within the given time period ConnectionPoolTimeoutException will be thrown.

One of the major shortcoming of the classic blocking I/O model is that the network socket can react to I/O events only when blocked in an I/O operation. When a connection is released back to the manager, it can be kept alive however it is unable to monitor the status of the socket and react to any I/O events. If the connection gets closed on the server side, the client side connection is unable to detect the change in the connection state and react appropriately by closing the socket on its end. HttpClient tries to mitigate the problem by testing whether the connection is 'stale', that is no longer valid because it was closed on the server side, prior to using the connection for executing an HTTP request. The stale connection check is not 100% reliable and adds 10 to 30 ms overhead to each request execution. The only feasible solution that does not involve a one thread per socket model for idle connections is a dedicated monitor thread used to evict connections that are considered expired due to a long period of inactivity. The monitor thread can periodically call ClientConnectionManager#closeExpiredConnections() method to close all expired connections and evict closed connections from the pool. It can also optionally call ClientConnectionManager#closeIdleConnections() method to close all connections that have been idle over a given period of time.

The HTTP specification does not specify how long a persistent connection may be and should be kept alive. Some HTTP servers use non-standard Keep-Alive header to communicate to the client the period of time in seconds they intend to keep the connection alive on the server side. HttpClient makes use of this information if available. If the Keep-Alive header is not present in the response, HttpClient assumes the connection can be kept alive indefinitely. However, many HTTP servers out there are configured to drop persistent connections after a certain period of inactivity in order to conserve system resources, quite often without informing the client. In case the default strategy turns out to be too optimistic, one may want to provide a custom keep-alive strategy.