=head1 NAME
Perlbal::Manual::Internals - Perlbal's architecture at a glance
=head2 VERSION
Perlbal 1.78.
=head2 DESCRIPTION
Connections come in from wherever and get to the TCPListener. It uses Service objects to determine what kind of Client* to spawn. The Client classes then handle crafting the response for the user.
{{ INTERNET }}
|
v
[Service]<===>[TCPListener]
___/ | \___
v v v
[ClientManage] [ClientHTTP] [ClientProxy]
^
|
v
[BackendHTTP]
Perlbal decides what backend to send a request to randomly (only presently supported method). If that service has idle backend connections available, configured by C<backend_persist> and C<connect_ahead>, it will reuse those connections and greatly reduce latency. See more detail in L<Perlbal::Manual::LoadBalancer>.
Perlbal also specializes in "spoonfeeding" data to slow clients. This allows backends to continue serving requests while Perlbal transfers responses back as fast as the client can read.
=head2 Classes
The following is a brief introduction/overview to the main Perlbal's classes:
=head3 Perlbal::Socket
Descends from L<Danga::Socket>.
Adds on to the base class to provide some functionality specifically useful for creating HTTP sockets.
=head4 Fields
=over 4
=item headers_string
Headers as they're being read.
=item req_headers
The final L<Perlbal::HTTPHeaders> object inbound.
=item res_headers
Response headers outbound (L<Perlbal::HTTPHeaders> object).
=item create_time
Creation time.
=item alive_time
Last time noted alive.
=item state
General purpose state; used by descendants.
=item do_die
If on, die and do no further requests.
=item read_buf
Arrayref of scalarref read from client.
=item read_ahead
Bytes sitting in read_buf.
=item read_size
Total bytes read from client, ever.
=item ditch_leading_rn
If true, the next header parsing will ignore a leading \r\n.
=item observed_ip_string
If defined, contains the observed IP string of the peer we're serving. This is intended for holding the value of the X-Forwarded-For and using it to govern ACLs.
=back
=head3 Perlbal::TCPListener
Descends from L<Perlbal::Socket>.
Very lightweight and fast connection accept class. Takes incoming connections as fast as possible and passes them off, instantiating one of the various Client* classes to handle it.
=head4 Fields
=over 4
=item service
L<Perlbal::Service>.
=item hostport
Scalar IP port of where this service is listening for new connections.
=item sslopts
The SSL Options.
use Data::Dumper;
warn Dumper( $tcp_listener->{'sslopts'} );
The above lines would print something like the following:
$VAR1 = {
'ssl' => {
'SSL_cipher_list' => '...',
'SSL_cert_file' => '...',
'SSL_key_file' => ',,,',
'SSL_ca_path' => '...',
'SSL_verify_mode' => '...'
}
};
=item v6
Boolean value stating whether the installation of Perlbal supports IPv6 (which basically boils down to L<Danga::Socket> v1.6.1 and L<IO::Socket::INET6> being available).
=back
=head3 Perlbal::BackendHTTP
Descends from L<Perlbal::Socket>.
This class handles connections to the backend web nodes for getting data back to the user. This class is used by other classes such as L<Perlbal::ClientProxy> to send a request to an internal node.
=head4 Fields
=over 4
=item client
L<Perlbal::ClientProxy> connection, or undef.
=item service
L<Perlbal::Service>.
=item pool
L<Perlbal::Pool>; whatever pool we spawned from.
=item ip
IP scalar.
=item port
Port scalar.
=item ipport
C<$ip:$port>.
=item reportto
Object; must implement reporter interface.
=item has_attention
Has been accepted by a webserver and we know for sure we're not just talking to the TCP stack.
=item waiting_options
If true, we're waiting for an OPTIONS * response to determine when we have attention.
=item disconnect_at
Time this connection will be disconnected, if it's kept-alive and backend told us; otherwise C<undef> for unknown.
=item content_length
Length of document being transferred. Only applies when the backend server sends a content-length header.
=item content_length_remain
Bytes remaining to be read. Only applies when the backend server sends a content-length header.
=item use_count
Number of requests this backend's been used for.
=item generation
Int; counts what generation we were spawned in.
=item buffered_upload_mode
Boolean. If on, we're doing a buffered upload transmit.
=item scratch
Extra storage; plugins can use it if they want.
=back
=head3 Perlbal::HTTPHeaders
Header management. Parses headers (request and response) and stores data for further user. Also manages validation of the request line so that it conforms to HTTP specifications.
=head4 Fields
=over 4
=item headers
href; lowercase header -> comma-sep list of values.
=item origcase
Href; lowercase header -> provided case.
=item hdorder
Aref; order headers were received (canonical order).
=item method
Scalar; request method (if GET request).
=item uri
Scalar; request URI (if GET request).
=item type
C<res> or C<req>.
=item code
HTTP response status code.
=item codetext
Status text that for response code.
=item ver
Version (string) "1.1".
=item vernum
Version (number: major*1000+minor): "1.1" => 1001.
=item responseLine
First line of HTTP response (if response).
=item requestLine
First line of HTTP request (if request).
=back
=head3 Perlbal::ClientHTTPBase
Descends from L<Perlbal::Socket>.
Provides base functionality to L<Perlbal::ClientHTTP> and L<Perlbal::ClientProxy>. Notably, the ability to efficiently send files to the remote user. Also handles most of the state logic for statistics and such. Is also used for services of type C<selector>. L<Perlbal::ClientHTTPBase> then reads in the request headers, and asks the service to re-bless the client instance to a more specific type, for either a L<Perlbal::ClientProxy> or L<Perlbal::ClientHTTP> (depending on selector's mapping).
=head4 Fields
=over 4
=item service
L<Perlbal::Service> object.
=item replacement_uri
URI to send instead of the one requested; this is used to instruct C<_serve_request> to send an index file instead of trying to serve a directory and failing.
=item scratch
Extra storage; plugins can use it if they want.
=item reproxy_file
Filename the backend told us to start opening.
=item reproxy_file_size
Size of file, once we C<stat()> it.
=item reproxy_fh
If needed, L<IO::Handle> of fd.
=item reproxy_file_offset
How much we've sent from the file.
=item post_sendfile_cb
Subref to run after we're done sendfile'ing the current file.
=item requests
Number of requests this object has performed for the user.
=item selector_svc
The original service from which we came.
=item is_ssl
Whether the socket was SSL attached (restricted operations).
=back
=head3 Perlbal::ClientHTTP
Descends from L<Perlbal::ClientHTTPBase>.
Very simple and lightweight class. Handles sending files to the user without much overhead. Most of the functionality is contained in the parent class, and this class doesn't implement much new stuff.
=head4 Fields
=over 4
=item put_in_progress
1 when we're currently waiting for an async job to return.
=item put_fh
File handle to use for writing data.
=item put_fh_filename
Filename of put_fh.
=item put_pos
File offset to write next data at.
=item content_length
Length of document being transferred.
=item content_length_remain
Bytes remaining to be read.
=item chunked_upload_state
Boolean/obj: if processing a chunked upload, L<Perlbal::ChunkedUploadState> object, else undef.
=back
=head3 Perlbal::ClientProxy
Descends from L<Perlbal::ClientHTTPBase>.
Takes an incoming connection from a user and connects to a backend node (C<Perlbal::BackendHTTP>) and relays the request. The backend can then either tell the proxy to reproxy and load a file from disk, or return a file directly, or just return a status message.
=head4 Fields
=over 4
=item backend
L<Perlbal::BackendHTTP> object (or C<undef> if disconnected).
=item backend_requested
True if we've requested a backend for this request.
=item reconnect_count
Number of times we've tried to reconnect to backend.
=item high_priority
Boolean; 1 if we are or were in the high priority queue.
=item low_priority
Boolean; 1 if we are or were in the low priority queue.
=item reproxy_uris
Arrayref; URIs to reproxy to, in order.
=item reproxy_expected_size
Int: size of response we expect to get back for reproxy.
=item currently_reproxying
Arrayref; the host info and URI we're reproxying right now.
=item content_length_remain
Int: amount of data we're still waiting for.
=item responded
Bool: whether we've already sent a response to the user or not.
=item last_request_time
Int: time that we last received a request.
=item primary_res_hdrs
If defined, we are doing a transparent reproxy-URI and the headers we get back aren't necessarily the ones we want. Instead, get most headers from the provided C<res> headers object here.
=item is_buffering
Bool; if we're buffering some/all of a request to memory/disk.
=item is_writing
Bool; if on, we currently have an C<aio_write> out.
=item start_time
Hi-res time when we started getting data to upload.
=item bufh
Buffered upload filehandle object.
=item bufilename
String; buffered upload filename.
=item bureason
String; if defined, the reason we're buffering to disk.
=item buoutpos
Int; buffered output position.
=item backend_stalled
Boolean: if backend has shut off its reads because we're too slow.
=item unread_data_waiting
Boolean: if we shut off reads while we know data is yet to be read from client.
=item chunked_upload_state
Bool/obj: if processing a chunked upload, L<Perlbal::ChunkedUploadState> object, else undef.
=item request_body_length
Integer: request's body length, either as-declared, or calculated after chunked upload is complete.
=item last_upload_packet
Unixtime we last sent a UDP upload packet. For perlbal sending out UDP packets related to upload status (for xmlhttprequest upload bar).
=item upload_session
Client's self-generated upload session. For perlbal sending out UDP packets related to upload status (for xmlhttprequest upload bar).
=item retry_count
Number of times we've retried this request so far after getting C<500> errors.
=back
=head3 Perlbal::ClientManage
Descends from L<Perlbal::Socket>.
Simple interface that provides a way for users to use the management interface of Perlbal. You can connect to the management port (as defined in the config file) with a web browser or regular telnet (see L<Perlbal::Manual::Management> for more information on this).
=head4 Fields
=over 4
=item service
L<Perlbal::Service>.
=item buf
Read buffer.
=item is_http
Boolean stating whether the request is HTTP.
=item ctx
L<Perlbal::CommandContext>.
=back
=head3 Perlbal::Service
A service is a particular item that Perlbal is doing. Services can have a role which defines how they behave. Each service can also have a bunch of parameters set to further adjust its behavior. By itself, the Service class handles maintaining pools of backend connections and managing statistics about itself.
=head4 Fields
=over 4
=item name
Name of the service.
=item role
Role type (C<web_server>, C<reverse_proxy>, etc).
=item enabled
Boolean; whether we're enabled or not (enabled = listening).
=item pool
L<Perlbal::Pool> that we're using to allocate nodes if we're in proxy mode.
=item listener
L<Perlbal::TCPListener> object, when enabled.
=item reproxy_cache
L<Perlbal::Cache> object, when enabled.
=back
=head4 End-user tunables
=over 4
=item listen
C<IP:port> of where we're listening for new connections.
=item docroot
Document root for C<web_server> role.
=item dirindexing
Boolean; directory indexing (for C<web_server> role). Not async.
=item index_files
Arrayref of filenames to try for index files.
=item enable_concatenate_get
Boolean; if user can request concatenated files.
=item enable_put
Boolean; whether PUT is supported.
=item max_put_size
Max size in bytes of a put file.
=item max_chunked_request_size
Max size in bytes of a chunked request (to be written to disk first).
=item min_put_directory
Number of directories required to exist at beginning of URIs in put.
=item enable_delete
Boolean; whether DELETE is supported.
=item high_priority_cookie
Cookie name to check if the client's requests should be considered high priority.
See also C<high_priority_cookie_contents>.
=item high_priority_cookie_contents
Aforementioned cookie value must contain this substring.
=item backend_persist_cache
Max number of persistent backends to hold onto while no clients.
=item persist_client
Boolean; persistent connections for clients.
=item persist_backend
Boolean; persistent connections for backends.
=item verify_backend
Boolean; get attention of backend before giving it clients (using OPTIONS).
=item verify_backend_path
Path to check with the OPTIONS request (default is C<*>).
=item max_backend_uses
Max requests to send per kept-alive backend (default 0 = unlimited).
=item connect_ahead
Number of spare backends to connect to in advance all the time.
=item buffer_size
How much data a L<Perlbal::ClientProxy> object should buffer from a backend.
=item buffer_size_reproxy_url
Same as above but for backends that are reproxying for us.
=item queue_relief_size
Number of outstanding standard priority connections to activate pressure relief at.
=item queue_relief_chance
Int, 0-100; % chance to take a standard priority request when we're in pressure relief mode.
=item trusted_upstream_proxies
Array of L<Net::Netmask> objects containing netmasks for trusted upstreams.
=item always_trusted
Boolean; if true, always trust upstreams.
=item blind_proxy
Boolean; if true, do not modify C<X-Forwarded-For>, C<X-Host>, or C<X-Forwarded-Host> headers.
=item enable_reproxy
Boolean; if true, advertise that server will reproxy files and/or URLs.
=item reproxy_cache_maxsize
Maximum number of reproxy results to be cached. (0 is disabled and default).
=item client_sndbuf_size
Bytes for C<SO_SNDBUF>.
=item server_process
Path to server process (executable).
=item persist_client_idle_timeout
Keep-alive timeout in seconds for clients (default is 30).
=item idle_timeout
Idle timeout outside of keep-alive time (default is 30).
=back
=head4 Internal state
=over 4
=item waiting_clients
Arrayref of clients waiting for backendhttp connections.
=item waiting_clients_highpri
Arrayref of high-priority clients waiting for backendhttp connections.
=item waiting_clients_lowpri
Arrayref of low-priority clients waiting for backendhttp connections.
=item waiting_client_count
Number of clients waiting for backends.
=item waiting_client_map
Map of clientproxy fd -> 1 (if they're waiting for a connection).
=item pending_connects
Hashref of C<ip:port> -> C<$time> (only one pending connect to backend at a time).
=item pending_connect_count
Number of outstanding backend connects.
=item bored_backends
Arrayref of backends we've already connected to, but haven't got clients.
=item hooks
Hashref: hookname => [ [ plugin, ref ], [ plugin, ref ], ... ].
=item plugins
Hashref: name => 1.
=item plugin_order
Arrayref: name, name, name...
=item plugin_setters
Hashref: { plugin_name => { key_name => coderef } }.
=item extra_config
Hashref with extra config options; name => values.
=item spawn_lock
Boolean; if true, we're currently in C<spawn_backends>.
=item extra_headers
{ insert => [ [ header, value ], ... ], remove => [ header, header, ... ], set => [ [ header, value ], ... ] }.
Used in header management interface.
=item generation
Int; generation count so we can slough off backends from old pools.
=item backend_no_spawn
{ "ip:port" => 1 }.
If on, C<spawn_backends> will ignore this C<ip:port> combo.
=item buffer_backend_connect
0 if off; otherwise, number of bytes to buffer before we ask for a backend.
=item selector
CODE ref, or undef, for role C<selector> services.
=item default_service
Name of a service a selector should default to.
=item buffer_uploads
Boolean; enable/disable the buffered uploads to disk system.
=item buffer_uploads_path
Path to store buffered upload files.
=item buffer_upload_threshold_time
Int; buffer uploads estimated to take longer than this.
=item buffer_upload_threshold_size
Int; buffer uploads greater than this size (in bytes).
=item buffer_upload_threshold_rate
Int; buffer uploads uploading at less than this rate (in bytes/sec).
=item upload_status_listeners
Comma separated list of C<ip:port> of UDP upload status receivers.
=item upload_status_listeners_sockaddr
Arrayref of sockaddrs (packed ip/port).
=item enable_ssl
Boolean; whether this service speaks SSL to the client.
=item ssl_key_file
File path to key pem file.
=item ssl_cert_file
File to path to cert pem file.
=item ssl_cipher_list
OpenSSL cipher list string.
=item ssl_ca_path
Path to certificates directory.
=item ssl_verify_mode
Int; verification mode, see L<IO::Socket::SSL>.
=item enable_error_retries
Boolean; whether we should retry requests after errors.
=item error_retry_schedule
Comma-separated seconds (full or partial) to delay between retries.
=item latency
Milliseconds of latency to add to request.
=item server_tokens
Boolean; whether to provide a "Server" header.
=item _stat_requests
Total requests to this service.
=item _stat_cache_hits
Total requests to this service that were served via the reproxy-url cache.
=back