FCGI-Spawn
view release on metacpan or search on metacpan
lib/FCGI/Spawn.pm view on Meta::CPAN
package FCGI::Spawn;
use vars qw($VERSION);
BEGIN {
$VERSION = '0.16.7';
$FCGI::Spawn::Default = 'FCGI::Spawn';
}
=pod
=head1 NAME
FCGI::Spawn - FastCGI server for CGI-like applications effective multiprocessing
=head1 SYNOPSIS
Minimum unrecommended way to illustrate it working:
FCGI::Spawn->new->spawn;
Never put this in production use. The C<fcgi_spawn> script supplied should care about sadly mandatory whistles and bells, at least the security is a king in sight of this:
FCGI::Spawn code should be run as its own user id, and the web server should be configured to request its FastCGI; in the case the local socket file is used, the web server should have the read and write permissions on it, the default name is /tmp/sp...
Consider about sock_chmod/sock_chown parameters for this, too.
In the case if you request via TCP care should be taken on network security like DMZ/VPN/firewalls setup instead of sock_* parameters.
About the ready to run applications compatibility refer to C<fcgi_spawn> docs.
Every other thing is explained in L<FCGI::ProcManager|FCGI::ProcManager> docs.
=head1 DESCRIPTION
The overall idea is to make Perl server-side scripts as convinient for newbies and server administrators as PHP in FastCGI mode.
FCGI::Spawn is used to serve as a FastCGI process manager.
Besides the features the FCGI::ProcManager posess itself, the FCGI::Spawn is targeted as web server admin understandable instance for building the own fastcgi server with copy-on-write memory sharing among forks and with single input parameters like...
Another thing to mention is that it is able to execute any file pointed by Web server ( FastCGI requester ).
So we have the daemon that is hot ready for hosting providing :-)
The definitive care is taken in FCGI::Spawn on security.
Besides the inode settings on local UNIX socket taken as input parameter, it is aware to avoid hosting users from changing the max_requests parameter meant correspondent to MaxRequests Apache's forked MPM parameter, and the respective current request...
The aforementioned max_requests parameter takes also care about the performance to avoid forks' memory leaks to consume all the RAM accounted on your hardware.
For shared hosting it is assumed that system administrator controls the process manager daemon script contents with those user hardware consumption limits and executes it with a user's credentials.
E. g., the user should be able to send signal to the daemon to initiate graceful restart on his/her demand ( this is yet to be done ) or change the settings those administrator can specifically allow in the daemon starter script without restart ( bot...
The call stack lets you set up your own code reference for your scripts execution. Also, feature exists that you can cache some object, like the template or XSLT processor and the recompilation to happen only on the template or xsl file(s) change. En...
Seeking for convention between high preformance needs that the perl compile cache possess and the convinience of debugging with recompilation on every request that php provides, the C<stats> feature allows you not to recompile the tested source like ...
This may be useful in both development environment to make the recompilation yet faster and on a production host to make the details of code adaptaion to hosting clear in a much less time needed.
=head1 Behind the scenes of fcgi_spawn
Here are the details for those unsatisfied with C<fcgi_spawn> but trying with FCGI::Spawn anyway:
=over
=item * You must configure socket
with %ENV or shell and tweak the CGI.pm with the patch supplied, if need, at early before use FCGI::Spawn.
%ENV and @INC can be tweaked in the special BEGIN block or you can eval "use FCGI::Spawn;" otherwise.
This is because of setting up the socket communication in the CGI::Fast, which is the part of Perl core distribution, right in the BEGIN block, e. g. right before the compilation.
=item * You should beware about CGI::Fast IS NOT included
at the moment this module is being used, e. g. IS ABSENT in the %INC global hash.
=item * CGI scripts ( if not CGI.pm ) must be tweaked to use $FCGI::Spawn::fcgi instead of new CGI or CGI->new.
In other case they will not be able to process HTTP POST.
In some caeses ( e. g. Bugzilla that derives CGI.pm ) the sources should be changed, too.
Hope your code obfuscators are not that complex to allow such a tweak. ;-)
FastCGI scripts do take the FastCGI object as a parameter, so it is obviously supplied in the callout code reference ( see below ).
=back
=head2 Why not mod_perl/mod_perl2/mod_fcgid?
lib/FCGI/Spawn.pm view on Meta::CPAN
=item * Too much need for root
The startup.pl providing the memory sharing among forks is aimed to be run as root, at least when you need to listen binded to ports numbered less than 1024, for example, 80.
And, the root user ( the human ) is often too busy to check if that massive code is secure enough to be run as root system user ( the effective UID ) Thus, it's no much deal to accelerate Perl on mod_perl steroids if the startup.pl includes rather sm...
Root is needed to recompile the Perl sources, at least with the useful Registry handler.
It is obvious to gracefully restart Apache once per N minutes and this is what several hosting panel use to do but it is not convinient to debug code that is migrated from developer's hosting to production's as it is needed to be recompiled on webma...
And, with no ( often proprietary ) hosting panel software onboard, Apache doesn't even gracefully restart on a regular basis without special admin care taken at server setup time.
On the uptime had gone till the need of restart after launch it is not an admin favor to do this, even gracefully.
Apache::Reload can save from this but it's not a production feature, too.
=item * File serving speed loss
Apache itself is well known to serve the static files too slowly.
Despite the promises of "we do this too" on sendfile and kqueue features.
=item * More stuff on your board overheads administering
The unclear differences between the bundled PerlHandler-s environments among with the more code to compile and depend on, causes the root changes are needed in the existing, mostly CGI, programs.
=back
=head2 Why still mod_perl?
=over
=item * the Habit
.
=item * Apache::DBI
Persistent connections feature makes it slightly faster for skip connect to DB stage on every request.
=item * Apache::Request, Apache::Session, etc.
HTTP input promises to be much more effective than the CGI.pm's one, used in CGI::Fast, too.
You may need also more information about request, like full incoming headers, too.
Those are obvious to be contained in the Apache internal structures rather than outsourced with application protocol from web server.
=back
=head2 Why not simply FCGI::ProcManager?
Targeted as a library for a single particular application or applications' framework, it takes you to make FCGI::Spawn and fcgi_spawn to obtain the ready to use application-agnostic daemon.
It seems to require too much in Perl knowledge from regular system administrator ( same as for startup.pl audit goes here ), in comparison to php's fastcgi mode.
Even with that, it is not as mock as FCGI::Spawn for software developer.
You will need to be me if you will need its features, if you are a sysadmin, while I'm the both.
=head1 PREREQUISITES
Be sure to have L<FCGI::ProcManager|FCGI::ProcManager>.
=head1 METHODS
=head2 new({hash parameters})
Class method.
Constructs a new process manager.
Takes an option hash of the sock_name and sock_chown initial parameter values, and passes the entire hash rest to ProcManager's constructor.
The parameters are:
=over
=item * $ENV{FCGI_SOCKET_PATH}
Not a hash parameter but the enironment variable.
Should be set before module compilation, to know where the socket resides.
Can be in the host:port or even :port notation for TCP, as FCGI.pm's remote.fpl states.
Default: /tmp/spawner.sock.
You can set environment value with your shell like this:
FCGI_SOCKET_PATH=/var/lib/fcgi.sock ./fcgi_spawn.pl <parameters>
or you can enclose it into the eval() like that:
$ENV{FCGI_SOCKET_PATH} = '/var/lib/fcgi.sock';
eval( "use FCGI::Spawn;" ); die $@ if $@;
=item * sock_chown
is the array reference which sets the parameters for chown() builtin on newly created socket, when needed.
Default: none.
=item * readchunk
is the buffer size for user's source reading in plsrc function.
Deprecated and will be removed in future versions.
Default: 4096.
=item * maxlength
is the maximumum user's file size for the same.
Default: 100000.
=item * max_requests
is the maximum requests number served by every separate fork.
Default: 20.
=item * stats
Whether to do or not to do the stat() on every file on the %INC to recompile on change by mean of removal from %INC.
Default: 1.
=item * stats_policy
Array reference that defines what kind of changes on the every module file stat()'s change to track and in what order.
Default: FCGI::Spawn::statnames_to_policy( 'mtime' ) ( statnames_to_policy() function is described below ).
=item * x_stats and x_stats_policy
Same as stats and stats_policy but for xinc() feature ( see below ).
=item * clean_inc_hash
when set to 1 points to clean out the requested via FastCGI file from %INC after every procesed request.
when set to 2 points to clean out after every procesed request the every %INC change that the FastCGI requested file did.
Default: 0.
=item * clean_main_space
when set to true points to clean out the %main:: changes ( unset the global variables ) at the same time.
Default: 0.
=item * clean_inc_subnamespace
Points which namespace, and beneath, should be cleaned in the moment between callouts ( e. g., Bugzilla, WebGUI, MyApp::MyClass etc., depending upon what is your applications name ).
when is a scalar, makes the %INC to clean if the begin of the key equals to it.
when is array reference, treats every element as a scalar and does the same for every item, just the same if it was the scalar itself
You can use :: namespace separator, as well as / as it is in the %INC, both MyApp::MyNS and MyApp/MyNS are pretty valid.
You can use full file names, like this: 'path/required_lib.pl' for this argument, too.
As of high-load systems it is strongly discouraged that the hosting user ( who can happen to be a really bad programmer ) to control this parameter, as it can lead to the same as clean_inc_hash=2 and can steal server performance at the moment just un...
ulimit is a good thing to keep from such a bothering too, but it's just not always sufficient alone. And, no ulimit on Cygwin by far.
Default: empty.
lib/FCGI/Spawn.pm view on Meta::CPAN
}
eval("use CGI::Fast;");
die $@ if $@;
}
my $readchunk = 4096;
# Deprecated and will be removed in future versions. And readchunk too.
sub plsrc {
my $sn = shift;
unless ( open PLSRC, $sn ) { exit $!; }
my $plsrc = "";
while ( my $rv = read( PLSRC, my $buf, $readchunk ) ) {
unless ( defined $rv ) { exit $!; }
$plsrc .= $buf;
exit if length($plsrc) > $maxlength;
}
close PLSRC;
return \$plsrc;
}
my $defaults = {
n_processes => 5,
max_requests => 20,
clean_inc_hash => 0,
clean_main_space => 0,
clean_inc_subnamespace => [],
callout => sub {
do shift;
},
stats => 1,
stats_policy => statnames_to_policy('mtime'),
x_stats => 1,
x_stats_policy => statnames_to_policy('mtime'),
state => {},
seed_rand => 1,
save_env => 1,
procname => 1,
is_prepared => 0,
};
sub statnames_to_policy {
my %policies
= qw/dev 0 ino 1 mode 2 nlink 3 uid 4 gid 5 rdev 6 size 7 atime 8 mtime 9 ctime 10 blksize 11 blocks 12/;
grep( { $_ eq 'all' } @_ )
? [ 0 .. 7, 9 .. 12 ]
: [ map( { $policies{$_} } @_ ) ];
}
sub new {
my $class = shift;
my ( $new_properties, $properties );
if ( $properties = shift ) {
$properties = { %$defaults, %$properties };
}
else {
$properties = $defaults;
}
my $proc_manager = FCGI::ProcManager->new($properties);
my $sock_name = $ENV{FCGI_SOCKET_PATH};
if ( defined $properties->{sock_chown} ) {
chown( @{ $properties->{sock_chown} }, $sock_name )
or die $!;
}
if ( defined $properties->{sock_chmod} ) {
chmod( $properties->{sock_chmod}, $sock_name )
or die $!;
}
defined $properties->{readchunk}
and $readchunk = $properties->{readchunk};
defined $properties->{maxlength}
and $maxlength = $properties->{maxlength};
$class->make_clean_inc_subnamespace($properties);
$properties->{proc_manager} = $proc_manager;
bless $properties, $class;
}
sub make_clean_inc_subnamespace {
my ( $self, $properties ) = @_;
my $cisns = $properties->{clean_inc_subnamespace};
if ( '' eq ref $cisns ) {
$cisns = [$cisns];
}
foreach (@$cisns) {
$_ =~ s!::!/!g
if '' eq ref $_;
}
$properties->{clean_inc_subnamespace} = $cisns;
}
sub _callout {
my $self = shift;
my %save_env = %ENV if $self->{save_env};
my $procname;
if ( $self->{procname} ) {
$procname = $0;
$0 = $_[0];
}
$self->{callout}->(@_);
$0 = $procname if $self->{procname};
%ENV = %save_env if $self->{save_env};
}
sub callout {
my $self = shift;
$self->_callout(@_);
$self->postspawn_dispatch;
}
sub clean_inc_particular {
my $self = shift;
map {
my $subnamespace_to_clean = $_;
map { delete $INC{$_} }
grep {
$subnamespace_to_clean eq substr $_, 0,
length $subnamespace_to_clean
}
keys %INC
} @{ $self->{clean_inc_subnamespace} };
( run in 1.675 second using v1.01-cache-2.11-cpan-71847e10f99 )