Provided by: pagekite_1.5.2.200603-3_all
NAME
pagekite - Make localhost servers publicly visible
SYNOPSIS
pagekite [--options] [service] kite-name [+flags]
DESCRIPTION
PageKite is a system for exposing localhost servers to the public Internet. It is most commonly used to make local web servers or SSH servers publicly visible, although almost any TCP-based protocol can work if the client knows how to use an HTTP proxy. PageKite uses a combination of tunnels and reverse proxies to compensate for the fact that localhost usually does not have a public IP address and is often subject to adverse network conditions, including aggressive firewalls and multiple layers of NAT. This program implements both ends of the tunnel: the local "back-end" and the remote "front-end" reverse-proxy relay. For convenience, pagekite also includes a basic HTTP server for quickly exposing files and directories to the World Wide Web for casual sharing and collaboration.
BASIC USAGE
Basic usage, gives http://localhost:80/ a public name: $ pagekite NAME.pagekite.me To expose specific folders, files or use alternate local ports: $ pagekite /a/path/ NAME.pagekite.me +indexes # built-in HTTPD $ pagekite *.html NAME.pagekite.me # built-in HTTPD $ pagekite 3000 NAME.pagekite.me # HTTPD on 3000 To expose multiple local servers (SSH and HTTP): $ pagekite ssh://NAME.pagekite.me AND 3000 NAME.pagekite.me
SERVICES AND KITES
The most comman usage of pagekite is as a back-end, where it is used to expose local services to the outside world. Examples of services are: a local HTTP server, a local SSH server, a folder or a file. A service is exposed by describing it on the command line, along with the desired public kite name. If a kite name is requested which does not already exist in the configuration file and program is run interactively, the user will be prompted and given the option of signing up and/or creating a new kite using the pagekite.net service. Multiple services and kites can be specified on a single command-line, separated by the word 'AND' (note capital letters are required). This may cause problems if you have many files and folders by that name, but that should be relatively rare. :-)
KITE CONFIGURATION
The options --list, --add, --disable and --remove can be used to manipulate the kites and service definitions in your configuration file, if you prefer not to edit it by hand. Examples: Adding new kites $ pagekite --add /a/path/ NAME.pagekite.me +indexes $ pagekite --add 80 OTHER-NAME.pagekite.me To display the current configuration $ pagekite --list Disable or delete kites (--add re-enables) $ pagekite --disable OTHER-NAME.pagekite.me $ pagekite --remove NAME.pagekite.me
FLAGS
Flags are used to tune the behavior of a particular kite, for example by enabling access controls or specific features of the built-in HTTP server. Common flags +ip/1.2.3.4 Enable connections only from this IP address. +ip/1.2.3 Enable connections only from this /24 netblock. HTTP protocol flags +password/name=pass Require a username and password (HTTP Basic Authentication) +rewritehost Rewrite the incoming Host: header. +rewritehost=N Replace Host: header value with N. +rawheaders Do not rewrite (or add) any HTTP headers at all. +proxyproto Use HAProxy's PROXY Protocol (v1) to relay IPs etc. +insecure Allow access to phpMyAdmin, /admin, etc. (per kite). Built-in HTTPD flags +indexes Enable directory indexes. +indexes=all Enable directory indexes including hidden (dot-) files. +hide Obfuscate URLs of shared files. +uploads Accept file uploads. +uploads=RE Accept uploads to paths matching regexp RE. +ul_filenames=PUpload naming policy. P = overwrite, keep or rename +cgi=list A list of extensions, for which files should be treated as CGI scripts (example: +cgi=cgi,pl,sh). +photobackup=password Enable built-in PhotoBackup server with the given password. See https://photobackup.github.io/ for details.
OPTIONS
The full power of pagekite lies in the numerous options which can be specified on the command line or in a configuration file (see below). Note that many options, especially the service and domain definitions, are additive and if given multiple options the program will attempt to obey them all. Options are processed in order and if they are not additive then the last option will override all preceding ones. Although pagekite accepts a great many options, most of the time the program defaults will Just Work. Common options --clean Skip loading the default configuration file. --signup Interactively sign up for pagekite.net service. --defaults Set defaults for use with pagekite.net service. --whitelabel=D Set defaults for pagekite.net white-labels. --whitelabels=D Set defaults for pagekite.net white-labels (with TLS). --nocrashreport Don't send anonymous crash reports to pagekite.net. Back-end options --shell Run PageKite in an interactive shell. --nullui Silent UI for scripting. Assumes Yes on all questions. --list List all configured kites. --add Add (or enable) the following kites, save config. --remove Remove the following kites, save config. --disable Disable the following kites, save config. --only Disable all but the following kites, save config. --insecure Allow access to phpMyAdmin, /admin, etc. (global). --local=ports Configure for local serving only (no remote front-end). --watch=N Display proxied data (higher N = more verbosity). --noproxy Ignore system (or config file) proxy settings. --proxy=type:server:port, --socksify=server:port, --torify=server:port Connect to the front-ends using SSL, an HTTP proxy, a SOCKS proxy, or the Tor anonymity network. The type can be any of 'ssl', 'http' or 'socks5'. The server name can either be a plain hostname, user@hostname or user:password@hostname. For SSL connections the user part may be a path to a client cert PEM file. If multiple proxies are defined, they will be chained one after another. --service_on=proto:kitename:host:port:secret Explicit configuration for a service kite. Generally kites are created on the command-line using the service short-hand described above, but this syntax is used in the config file. The kitename `unknown`, if allowed by the front-end, represents a backend of last resort for requests with no other match. --authdomain=DNS-suffix, --authdomain=/path/to/app, --authdomain=kite-domain:DNS-suffix, --authdomain=kite-domain:/path/to/app Use DNS-suffix for remote DNS-based authentication of incoming tunnel requests, or invoke an external application for this purpose. If no kite-domain is given, use this as the default authentication method. See the section below on tunnel authentication for further details. In order for the app path to be recognized as such, it must contain at least one / character. --auththreads=N Start N threads to process auth requests. Default is 1. --authfail_closed If authentication fails, reject tunnel requests. The default is to fail open and allow tunnels if the auth checks are broken. --service_off=proto:kitename:host:port:secret Same as --service_on, except disabled by default. --service_cfg=..., --webpath=... These options are used in the configuration file to store service and flag settings (see above). These are both likely to change in the near future, so please just pretend you didn't notice them. --frontend=host:port Connect to the named front-end server. If this option is repeated, multiple connections will be made. --frontends=num:dns-name:port Choose num front-ends from the A records of a DNS domain name, using the given port number. Default behavior is to probe all addresses and use the fastest one. --frontends=num:@/path/to/file:port Same as above, except the IP address list will be loaded from a file (and reloaded periodically), instead of using DNS. --nofrontend=ip:port Never connect to the named front-end server. This can be used to exclude some front-ends from auto-configuration. --fe_certname=domain Connect using SSL, accepting valid certs for this domain. If this option is repeated, any of the named certificates will be accepted, but the first will be preferred. --fe_nocertcheck Connect using SSL/TLS, but do not verify the remote certificate. This is largely insecure but still thwarts passive attacks and prevents routers and firewalls from corrupting the PageKite tunnel. --ca_certs=/path/to/file Path to your trusted root SSL certificates file. --dyndns=X Register changes with DynDNS provider X. X can either be simply the name of one of the 'built-in' providers, or a URL format string for ad-hoc updating. --keepalive=N Force traffic over idle tunnels every N seconds, to cope with firewalls that kill idle TCP connections. Backend only: if set to "auto" (the default), the interval will be adjusted automatically in response to disconnects. --all Terminate early if any tunnels fail to register. --new Don't attempt to connect to any kites' old front-ends. --noprobes Reject all probes for service state. Front-end options --isfrontend Enable front-end operation. --domain=proto,proto2,pN:domain:secret Accept tunneling requests for the named protocols and specified domain, using the given secret. A * may be used as a wildcard for subdomains or protocols. This is for static configurations, for dynamic access controls use the `--authdomain` mechanism. The domain `unknown`, if configured, represents a backend of last resort for incoming requests with no other match. --authdomain=DNS-suffix, --authdomain=/path/to/app, --authdomain=kite-domain:DNS-suffix, --authdomain=kite-domain:/path/to/app Use DNS-suffix for remote DNS-based authentication of incoming tunnel requests, or invoke an external application for this purpose. If no kite-domain is given, use this as the default authentication method. See the section below on tunnel authentication for further details. In order for the app path to be recognized as such, it must contain at least one / character. --auththreads=N Start N threads to process auth requests. Default is 1. --authfail_closed If authentication fails, reject tunnel requests. The default is to fail open and allow tunnels if the auth checks are broken. --motd=/path/to/motd Send the contents of this file to new back-ends as a "message of the day". --host=hostnameListen on the given hostname only. --ports=list Listen on a comma-separated list of ports. --portalias=A:BReport port A as port B to backends (because firewalls). --protos=list Accept the listed protocols for tunneling. --rawports=list Listen for raw connections these ports. The string '%s' allows arbitrary ports in HTTP CONNECT. --overload=baseline, --overload_cpu=fraction, 0-1, --overload_mem=fraction, 0-1 Enable "overload" calculations, which cause the front-end to recommend back-ends go elsewhere if possible, once connection counts go above a certain number. The baseline is the initial overload level, but it will be adjusted dynamically based on load average (CPU use) and memory usage. This will really only work well on Linux and if PageKite is the only thing happening on the machine. Setting both fractions to 0 disables dynamic scaling. --overload_file=/path/to/baseline/file Path to a file, the contents of which overrides all overload calculations. This can be used to manage load calculations using an external process (or by hand, e.g. to prepare for maintenance). Note that overload must specify a non-zero baseline, otherwise this setting is ignored. --ratelimit_ips=IPs/seconds, --ratelimit_ips=kitename:IPs/seconds Limit how many different clients (IPs) can request data from a tunnel within a given window of time, e.g. 5/3600. This is useful as either a crude form of DDoS mitigation, or as a mechanism to make public kite services unusable for phishing. Note that limits are enforced per-tunnel (not per kite), and tunnels serving multiple kites will use the settings of the strictest kite. Limits apply to subdomains as well. A single IP may be counted more than once if request headers (such as User-Agent) differ. --accept_acl_file=/path/to/file Consult an external access control file before accepting an incoming connection. Quick'n'dirty for mitigating abuse. The format is one rule per line: `rule policy comment` where a rule is an IP or regexp and policy is 'allow' or 'deny'. --client_acl=policy:regexp, --tunnel_acl=policy:regexp Add a client connection or tunnel access control rule. Policies should be 'allow' or 'deny', the regular expression should be written to match IPv4 or IPv6 addresses. If defined, access rules are checkd in order and if none matches, incoming connections will be rejected. --tls_default=name Default name to use for SSL, if SNI (Server Name Indication) is missing from incoming HTTPS connections. --tls_endpoint=name:/path/to/file Terminate SSL/TLS for a name using key/cert from a file. System options --optfile=/path/to/file Read settings from file X. Default is ~/.pagekite.rc. --optdir=/path/to/directory Read settings from /path/to/directory/*.rc, in lexicographical order. --savefile=/path/to/file Saved settings will be written to this file. --save Save the current configuration to the savefile. --settings Dump the current settings to STDOUT, formatted as a configuration file would be. --nopyopenssl Avoid use of the pyOpenSSL library (not in config file) --nossl Avoid use SSL entirely (not allowed in config file) --nozchunks Disable zlib tunnel compression. --sslzlib Enable zlib compression in OpenSSL. --buffers=N Buffer at most N kB of data before blocking. --logfile=F Log to file F, stdio means standard output. --daemonize Run as a daemon. --runas=U:G Set UID:GID after opening our listening sockets. --pidfile=P Write PID to the named file. --errorurl=U URL to redirect to when back-ends are not found. --errorurl=D:UCustom error URL for domain D. --selfsign Configure the built-in HTTP daemon for HTTPS, first generating a new self-signed certificate using openssl if necessary. --httpd=X:P, --httppass=X, --pemfile=X Configure the built-in HTTP daemon. These options are likely to change in the near future, please pretend you didn't see them.
CONFIGURATION FILES
If you are using pagekite as a command-line utility, it will load its configuration from a file in your home directory. The file is named .pagekite.rc on Unix systems (including Mac OS X), or pagekite.cfg on Windows. If you are using pagekite as a system-daemon which starts up when your computer boots, it is generally configured to load settings from /etc/pagekite.d/*.rc (in lexicographical order). In both cases, the configuration files contain one or more of the same options as are used on the command line, with the difference that at most one option may be present on each line, and the parser is more tolerant of white-space. The leading '--' may also be omitted for readability and blank lines and lines beginning with '#' are treated as comments. NOTE: When using -o, --optfile or --optdir on the command line, it is advisable to use --clean to suppress the default configuration.
SECURITY
Please keep in mind, that whenever exposing a server to the public Internet, it is important to think about security. Hacked webservers are frequently abused as part of virus, spam or phishing campaigns and in some cases security breaches can compromise the entire operating system. Some advice: * Switch PageKite off when not using it. * Use the built-in access controls and SSL encryption. * Leave the firewall enabled unless you have good reason not to. * Make sure you use good passwords everywhere. * Static content is very hard to hack! * Always, always make frequent backups of any important work. Note that as of version 0.5, pagekite includes a very basic request firewall, which attempts to prevent access to phpMyAdmin and other sensitive systems. If it gets in your way, the +insecure flag or --insecure option can be used to turn it off. For more, please visit: <https://pagekite.net/support/security/>
TUNNEL REQUEST AUTHENTICATION
When running pagekite as a front-end relay, you can enable dynamic authentication of incoming tunnel requests in two ways. One uses a DNS-based protocol for delegating authentication to a remote server. The nice thing about this, is relays can be deployed without any direct access to your user account databases - in particular, a zero-knowlege challenge/response protocol is used which means the relay never sees the shared secret used to authenticate the kite. The second method delegates authentication to an external app; this external app can be written in any language you like, as long as it implements the following command-line arguments: --capabilities Print a list of capabilities to STDOUT and exit --server Run as a "server", reading queries on STDIN and sending one-line replies to STDOUT. --auth <domain> Return JSON formatted auth and quota details --zk-auth <query> Implement the DNS-based zero-knowlege protocol The recognized capabilities are SERVER, ZK-AUTH and AUTH. One of AUTH or ZK-AUTH is required. The JSON `--auth` responses should be dictionaries which have at least one element, `secret` or `error`. The secret is the shared secret to be used to authenticate the tunnel. The dictionary may also contain advisory quota values (`quota_kb`, `quota_days` and `quota_conns`), and IP rate limiting parameters (`ips_per_sec-ips` and `ips_per_sec-secs`). The source distribution of pagekite includes a script named `demo_auth_app.py` which implements this protocol.
BUGS
Using pagekite as a front-end relay with the native Python SSL module may result in poor performance. Please use the pyOpenSSL wrappers instead.
SEE ALSO
lapcat(1), <http://pagekite.org/>, <https://pagekite.net/>
CREDITS
- Bjarni R. Einarsson <http://bre.klaki.net/> - The Beanstalks Project ehf. <https://pagekite.net/company/> - The Rannis Technology Development Fund <http://www.rannis.is/> - Joar Wandborg <http://wandborg.se/> - Luc-Pierre Terral
COPYRIGHT AND LICENSE
Copyright 2010-2020, the Beanstalks Project ehf. and Bjarni R. Einarsson. This program is free software: you can redistribute it and/or modify it under the terms of the GNU Affero General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Affero General Public License for more details. You should have received a copy of the GNU Affero General Public License along with this program. If not, see: <http://www.gnu.org/licenses/>