diff options
Diffstat (limited to 'actionpack/lib/action_dispatch')
| -rw-r--r-- | actionpack/lib/action_dispatch/middleware/remote_ip.rb | 165 |
1 files changed, 112 insertions, 53 deletions
diff --git a/actionpack/lib/action_dispatch/middleware/remote_ip.rb b/actionpack/lib/action_dispatch/middleware/remote_ip.rb index 5abf8f2802..ad5e1993aa 100644 --- a/actionpack/lib/action_dispatch/middleware/remote_ip.rb +++ b/actionpack/lib/action_dispatch/middleware/remote_ip.rb @@ -1,23 +1,58 @@ module ActionDispatch + # This middleware calculates the IP address of the remote client that is + # making the request. It does this by checking various headers that could + # contain the address, and then picking the last-set address that is not + # on the list of trusted IPs. This follows the precendent set by e.g. + # {the Tomcat server}[https://issues.apache.org/bugzilla/show_bug.cgi?id=50453], + # with {reasoning explained at length}[http://blog.gingerlime.com/2012/rails-ip-spoofing-vulnerabilities-and-protection] + # by @gingerlime. A more detailed explanation of the algorithm is given + # at GetIp#calculate_ip. + # + # Some Rack servers concatenate repeated headers, like {HTTP RFC 2616}[http://www.w3.org/Protocols/rfc2616/rfc2616-sec4.html#sec4.2] + # requires. Some Rack servers simply drop preceeding headers, and only report + # the value that was {given in the last header}[http://andre.arko.net/2011/12/26/repeated-headers-and-ruby-web-servers]. + # If you are behind multiple proxy servers (like Nginx to HAProxy to Unicorn) + # then you should test your Rack server to make sure your data is good. + # + # IF YOU DON'T USE A PROXY, THIS MAKES YOU VULNERABLE TO IP SPOOFING. + # This middleware assumes that there is at least one proxy sitting around + # and setting headers with the client's remote IP address. If you don't use + # a proxy, because you are hosted on e.g. Heroku, any client can claim to + # have any IP address by setting the X-Forwarded-For header. If you care + # about that, please take precautions. class RemoteIp - class IpSpoofAttackError < StandardError ; end + class IpSpoofAttackError < StandardError; end - # IP addresses that are "trusted proxies" that can be stripped from - # the comma-delimited list in the X-Forwarded-For header. See also: - # http://en.wikipedia.org/wiki/Private_network#Private_IPv4_address_spaces - # http://en.wikipedia.org/wiki/Private_network#Private_IPv6_addresses. + # The default trusted IPs list simply includes IP addresses that are + # guaranteed by the IP specification to be private addresses. Those will + # not be the ultimate client IP in production, and so are discarded. See + # http://en.wikipedia.org/wiki/Private_network for details. TRUSTED_PROXIES = %r{ - ^127\.0\.0\.1$ | # localhost - ^::1$ | - ^(10 | # private IP 10.x.x.x - 172\.(1[6-9]|2[0-9]|3[0-1]) | # private IP in the range 172.16.0.0 .. 172.31.255.255 - 192\.168 | # private IP 192.168.x.x - fc00:: # private IP fc00 - )\. + ^127\.0\.0\.1$ | # localhost IPv4 + ^::1$ | # localhost IPv6 + ^fc00: | # private IPv6 range fc00 + ^10\. | # private IPv4 range 10.x.x.x + ^172\.(1[6-9]|2[0-9]|3[0-1])\.| # private IPv4 range 172.16.0.0 .. 172.31.255.255 + ^192\.168\. # private IPv4 range 192.168.x.x }x attr_reader :check_ip, :proxies + # Create a new +RemoteIp+ middleware instance. + # + # The +check_ip_spoofing+ option is on by default. When on, an exception + # is raised if it looks like the client is trying to lie about its own IP + # address. It makes sense to turn off this check on sites aimed at non-IP + # clients (like WAP devices), or behind proxies that set headers in an + # incorrect or confusing way (like AWS ELB). + # + # The +custom_trusted+ argument can take a regex, which will be used + # instead of +TRUSTED_PROXIES+, or a string, which will be used in addition + # to +TRUSTED_PROXIES+. Any proxy setup will put the value you want in the + # middle (or at the beginning) of the X-Forwarded-For list, with your proxy + # servers after it. If your proxies aren't removed, pass them in via the + # +custom_trusted+ parameter. That way, the middleware will ignore those + # IP addresses, and return the one that you want. def initialize(app, check_ip_spoofing = true, custom_proxies = nil) @app = app @check_ip = check_ip_spoofing @@ -31,15 +66,23 @@ module ActionDispatch end end + # Since the IP address may not be needed, we store the object here + # without calculating the IP to keep from slowing down the majority of + # requests. For those requests that do need to know the IP, the + # GetIp#calculate_ip method will calculate the memoized client IP address. def call(env) env["action_dispatch.remote_ip"] = GetIp.new(env, self) @app.call(env) end + # The GetIp class exists as a way to defer processing of the request data + # into an actual IP address. If the ActionDispatch::Request#remote_ip method + # is called, this class will calculate the value and then memoize it. class GetIp - # IP v4 and v6 (with compression) validation regexp - # https://gist.github.com/1289635 + # This constant contains a regular expression that validates every known + # form of IP v4 and v6 address, with or without abbreviations, adapted + # from {this gist}[https://gist.github.com/1289635]. VALID_IP = %r{ (^(25[0-5]|2[0-4][0-9]|1[0-9][0-9]|[0-9]{1,2})(\.(25[0-5]|2[0-4][0-9]|1[0-9][0-9]|[0-9]{1,2})){3}$) | # ip v4 (^( @@ -63,62 +106,78 @@ module ActionDispatch }x def initialize(env, middleware) - @env = env - @middleware = middleware - @ip = nil + @env = env + @check_ip = middleware.check_ip + @proxies = middleware.proxies end - # Determines originating IP address. REMOTE_ADDR is the standard - # but will be wrong if the user is behind a proxy. Proxies will set - # HTTP_CLIENT_IP and/or HTTP_X_FORWARDED_FOR, so we prioritize those. - # HTTP_X_FORWARDED_FOR may be a comma-delimited list in the case of - # multiple chained proxies. The first address which is in this list - # if it's not a known proxy will be the originating IP. - # Format of HTTP_X_FORWARDED_FOR: - # client_ip, proxy_ip1, proxy_ip2... - # http://en.wikipedia.org/wiki/X-Forwarded-For + # Sort through the various IP address headers, looking for the IP most + # likely to be the address of the actual remote client making this + # request. + # + # REMOTE_ADDR will be correct if the request is made directly against the + # Ruby process, on e.g. Heroku. When the request is proxied by another + # server like HAProxy or Nginx, the IP address that made the original + # request will be put in an X-Forwarded-For header. If there are multiple + # proxies, that header may contain a list of IPs. Other proxy services + # set the Client-Ip header instead, so we check that too. + # + # As discussed in {this post about Rails IP Spoofing}[http://blog.gingerlime.com/2012/rails-ip-spoofing-vulnerabilities-and-protection/], + # while the first IP in the list is likely to be the "originating" IP, + # it could also have been set by the client maliciously. + # + # In order to find the first address that is (probably) accurate, we + # take the list of IPs, remove known and trusted proxies, and then take + # the last address left, which was presumably set by one of those proxies. def calculate_ip - client_ip = @env['HTTP_CLIENT_IP'] - forwarded_ip = ips_from('HTTP_X_FORWARDED_FOR').first - remote_addrs = ips_from('REMOTE_ADDR') - - check_ip = client_ip && @middleware.check_ip - if check_ip && forwarded_ip != client_ip + # Set by the Rack web server, this is a single value. + remote_addr = ips_from('REMOTE_ADDR').last + + # Could be a CSV list and/or repeated headers that were concatenated. + client_ips = ips_from('HTTP_CLIENT_IP').reverse + forwarded_ips = ips_from('HTTP_X_FORWARDED_FOR').reverse + + # +Client-Ip+ and +X-Forwarded-For+ should not, generally, both be set. + # If they are both set, it means that this request passed through two + # proxies with incompatible IP header conventions, and there is no way + # for us to determine which header is the right one after the fact. + # Since we have no idea, we give up and explode. + should_check_ip = @check_ip && client_ips.last + if should_check_ip && !forwarded_ips.include?(client_ips.last) # We don't know which came from the proxy, and which from the user - raise IpSpoofAttackError, "IP spoofing attack?!" \ - "HTTP_CLIENT_IP=#{@env['HTTP_CLIENT_IP'].inspect}" \ + raise IpSpoofAttackError, "IP spoofing attack?! " + + "HTTP_CLIENT_IP=#{@env['HTTP_CLIENT_IP'].inspect} " + "HTTP_X_FORWARDED_FOR=#{@env['HTTP_X_FORWARDED_FOR'].inspect}" end - client_ips = remove_proxies [client_ip, forwarded_ip, remote_addrs].flatten - if client_ips.present? - client_ips.first - else - # If there is no client ip we can return first valid proxy ip from REMOTE_ADDR - remote_addrs.find { |ip| valid_ip? ip } - end + # We assume these things about the IP headers: + # + # - X-Forwarded-For will be a list of IPs, one per proxy, or blank + # - Client-Ip is propagated from the outermost proxy, or is blank + # - REMOTE_ADDR will be the IP that made the request to Rack + ips = [forwarded_ips, client_ips, remote_addr].flatten.compact + + # If every single IP option is in the trusted list, just return REMOTE_ADDR + filter_proxies(ips).first || remote_addr end + # Memoizes the value returned by #calculate_ip and returns it for + # ActionDispatch::Request to use. def to_s @ip ||= calculate_ip end - private + protected def ips_from(header) - @env[header] ? @env[header].strip.split(/[,\s]+/) : [] - end - - def valid_ip?(ip) - ip =~ VALID_IP - end - - def not_a_proxy?(ip) - ip !~ @middleware.proxies + # Split the comma-separated list into an array of strings + ips = @env[header] ? @env[header].strip.split(/[,\s]+/) : [] + # Only return IPs that are valid according to the regex + ips.select{ |ip| ip =~ VALID_IP } end - def remove_proxies(ips) - ips.select { |ip| valid_ip?(ip) && not_a_proxy?(ip) } + def filter_proxies(ips) + ips.reject { |ip| ip =~ @proxies } end end |