=head1 NAME AnyEvent::SNMP - adaptor to integrate Net::SNMP into AnyEvent. =head1 SYNOPSIS use AnyEvent::SNMP; use Net::SNMP; # just use Net::SNMP and AnyEvent as you like: # use a condvar to transfer results, this is # just an example, you can use a naked callback as well. my $cv = AnyEvent->condvar; # ... start non-blocking snmp request(s)... Net::SNMP->session (-hostname => "127.0.0.1", -community => "public", -nonblocking => 1) ->get_request (-callback => sub { $cv->send (@_) }); # ... do something else until the result is required my @result = $cv->wait; =head1 DESCRIPTION This module implements an alternative "event dispatcher" for Net::SNMP, using AnyEvent as a backend. This integrates Net::SNMP into AnyEvent. That means you can make non-blocking Net::SNMP calls and as long as other parts of your program also use AnyEvent (or some event loop supported by AnyEvent), they will run in parallel. Also, the Net::SNMP scheduler is very inefficient with respect to both CPU and memory usage. Most AnyEvent backends (including the pure-perl backend) fare much better than the Net::SNMP dispatcher. Another major added feature of this module over Net::SNMP is automatic rate-adjustments: Net::SNMP is so slow that firing a few thousand requests can cause many timeouts simply because Net::SNMP cannot process the replies in time. This module automatically adapts the send rate to avoid false timeouts caused by slow reply processing. A potential disadvantage of this module is that replacing the dispatcher is not at all a documented thing to do, so future changes in Net::SNMP might break this module (or the many similar ones). This module does not export anything and does not require you to do anything special apart from loading it I. It is recommended but not required to load this module before C. =head1 GLOBAL VARIABLES =over 4 =item $AnyEvent::SNMP::MAX_OUTSTANDING (default: C<50>, dynamic) =item AnyEvent::SNMP::set_max_outstanding $new_value Use this package variable to restrict the number of outstanding SNMP requests at any point in time. Net::SNMP is very fast at creating and sending SNMP requests, but much slower at parsing (big, bulk) responses. This makes it easy to request a lot of data that can take many seconds to parse. In the best case, this can lead to unnecessary delays (and even time-outs, as the data has been received but not yet processed) and in the worst case, this can lead to packet loss, when the receive queue overflows and the kernel can no longer accept new packets. To avoid this, you can (and should) limit the number of outstanding requests to a number low enough so that parsing time doesn't introduce noticeable delays. Unfortunately, this number depends not only on processing speed and load of the machine running Net::SNMP, but also on the network latency and the speed of your SNMP agents. AnyEvent::SNMP tries to dynamically adjust this number upwards and downwards. Increasing C<$MAX_OUTSTANDING> will not automatically use the extra request slots. To increase C<$MAX_OUTSTANDING> and make C make use of the extra parallelity, call C with the new value, e.g.: AnyEvent::SNMP::set_max_outstanding 500; Although due to the dynamic adjustment, this might have little lasting effect. Note that you can use L to speed up parsing of responses considerably. =item $AnyEvent::SNMP::MIN_RECVQUEUE (default: C<8>) =item $AnyEvent::SNMP::MAX_RECVQUEUE (default: C<64>) These values specify the minimum and maximum receive queue length (in units of one response packet). When AnyEvent::SNMP handles $MAX_RECVQUEUE or more packets per iteration it will reduce $MAX_OUTSTANDING. If it handles less than $MIN_RECVQUEUE, it increases $MAX_OUTSTANDING. This has the result of adjusting the number of outstanding requests so that the recv queue is between the minimum and maximum, usually. This algorithm works reasonably well as long as the responses, response latencies and processing times are the same per packet on average. =back =head1 COMPATIBILITY This module may be used as a drop in replacement for the Net::SNMP::Dispatcher in existing programs. You can still call C to start the event-loop, but then you loose the benefit of mixing Net::SNMP events with other events. use AnyEvent::SNMP; use Net::SNMP; # just use Net::SNMP as before # ... start non-blocking snmp request(s)... Net::SNMP->session ( -hostname => "127.0.0.1", -community => "public", -nonblocking => 1, )->get_request (-callback => sub { ... }); snmp_dispatcher; =cut package AnyEvent::SNMP; use common::sense; # it is possible to do this without loading # Net::SNMP::Dispatcher, but much more awkward. use Net::SNMP::Dispatcher; # we could inherit fro Net:SNMP::Dispatcher, but since this is undocumented, # I'd rather see it die (and reported) than silenty and subtly fail. *msg_handle_alloc = \&Net::SNMP::Dispatcher::msg_handle_alloc; sub Net::SNMP::Dispatcher::instance { AnyEvent::SNMP:: } use Net::SNMP (); use AnyEvent (); our $VERSION = '6.02'; $Net::SNMP::DISPATCHER = instance Net::SNMP::Dispatcher; our $MESSAGE_PROCESSING = $Net::SNMP::Dispatcher::MESSAGE_PROCESSING; our $BUSY; our $DONE; # finished all jobs our @TRANSPORT; # fileno => [count, watcher] our @QUEUE; our $MAX_OUTSTANDING = 50; our $MIN_RECVQUEUE = 8; our $MAX_RECVQUEUE = 64; sub kick_job; sub _send_pdu { my ($pdu, $retries) = @_; # mostly copied from Net::SNMP::Dispatch # Pass the PDU to Message Processing so that it can # create the new outgoing message. my $msg = $MESSAGE_PROCESSING->prepare_outgoing_msg ($pdu); if (!defined $msg) { --$BUSY; kick_job; # Inform the command generator about the Message Processing error. $pdu->status_information ($MESSAGE_PROCESSING->error); return; } # Actually send the message. if (!defined $msg->send) { $MESSAGE_PROCESSING->msg_handle_delete ($pdu->msg_id) if $pdu->expect_response; # A crude attempt to recover from temporary failures. if ($retries-- > 0 && ($!{EAGAIN} || $!{EWOULDBLOCK} || $!{ENOSPC})) { my $retry_w; $retry_w = AE::timer $pdu->timeout, 0, sub { undef $retry_w; _send_pdu ($pdu, $retries); }; } else { --$BUSY; kick_job; } # Inform the command generator about the send() error. $pdu->status_information ($msg->error); return; } # Schedule the timeout handler if the message expects a response. if ($pdu->expect_response) { my $transport = $msg->transport; my $fileno = $transport->fileno; # register the transport unless ($TRANSPORT[$fileno][0]++) { $TRANSPORT[$fileno][1] = AE::io $transport->socket, 0, sub { for my $count (1..$MAX_RECVQUEUE) { # handle up to this many requests in one go # Create a new Message object to receive the response my ($msg, $error) = Net::SNMP::Message->new (-transport => $transport); if (!defined $msg) { die sprintf 'Failed to create Message object [%s]', $error; } # Read the message from the Transport Layer if (!defined $msg->recv) { if ($transport->connectionless) { # if we handled very few replies and we have queued work, try # to increase the parallelity as we probably can handle more. if ($count < $MIN_RECVQUEUE && @QUEUE) { ++$MAX_OUTSTANDING; kick_job; } } else { # for some reason, connected-oriented transports seem to need this delete $TRANSPORT[$fileno] unless --$TRANSPORT[$fileno][0]; } $msg->error; return; } # For connection-oriented Transport Domains, it is possible to # "recv" an empty buffer if reassembly is required. if (!$msg->length) { return; } # Hand the message over to Message Processing. if (!defined $MESSAGE_PROCESSING->prepare_data_elements ($msg)) { $MESSAGE_PROCESSING->error; return; } # Set the error if applicable. $msg->error ($MESSAGE_PROCESSING->error) if $MESSAGE_PROCESSING->error; # Notify the command generator to process the response. # Net::SNMP calls process_response_pdu, which simply calls callback_execute, # but some errors cause $msg to be of type Net::SNMP::Message, not Net::SMMP::PDU, # so we call the underlying callback_execute method which exists on both and # seems to do the right thing. $msg->callback_execute; # Cancel the timeout. my $rtimeout_w = $msg->timeout_id; if ($$rtimeout_w) { undef $$rtimeout_w; --$BUSY; kick_job; unless (--$TRANSPORT[$fileno][0]) { delete $TRANSPORT[$fileno]; return; } } } # when we end up here, we successfully handled $MAX_RECVQUEUE # replies in one iteration, so assume we are overloaded # and reduce the amount of parallelity. $MAX_OUTSTANDING = (int $MAX_OUTSTANDING * 0.95) || 1; }; } $msg->timeout_id (\(my $rtimeout_w = AE::timer $pdu->timeout, 0, sub { my $rtimeout_w = $msg->timeout_id; if ($$rtimeout_w) { undef $$rtimeout_w; delete $TRANSPORT[$fileno] unless --$TRANSPORT[$fileno][0]; } if ($retries--) { _send_pdu ($pdu, $retries); } else { $MESSAGE_PROCESSING->msg_handle_delete ($pdu->msg_id); $pdu->status_information ("No response from remote host '%s'", $pdu->hostname); --$BUSY; kick_job; } }) ); } else { --$BUSY; kick_job; } } sub kick_job { while ($BUSY < $MAX_OUTSTANDING) { my $pdu = shift @QUEUE or last; ++$BUSY; _send_pdu $pdu, $pdu->retries; } $DONE and $DONE->() unless $BUSY; } sub send_pdu($$$) { my (undef, $pdu, $delay) = @_; # $delay is not very sensibly implemented by AnyEvent::SNMP, # but apparently it is not a very sensible feature. if ($delay > 0) { ++$BUSY; my $delay_w; $delay_w = AE::timer $delay, 0, sub { undef $delay_w; push @QUEUE, $pdu; --$BUSY; kick_job; }; return 1; } push @QUEUE, $pdu; kick_job; 1 } sub loop($) { while ($BUSY) { $DONE = AE::cv; $DONE->recv; undef $DONE; } } *activate = \&loop; # 5.x compatibility? *listen = \&loop; # 5.x compatibility? sub one_event($) { # should not ever be used AnyEvent->one_event; #d# todo } sub set_max_outstanding($) { $MAX_OUTSTANDING = $_[0]; kick_job; } # not provided yet: # schedule # apparently only used by Net::SNMP::Dispatcher itself # register # apparently only used by Net::SNMP::Dispatcher itself # deregister # apparently only used by Net::SNMP::Dispatcher itself # cancel # apparently only used by Net::SNMP::Dispatcher itself # return_response_pdu # apparently not used at all? # error # only used by Net::SNMP::Dispatcher itself? # debug # only used by Net::SNMP::Dispatcher itself? =head1 SEE ALSO L, L, L, L. =head1 AUTHOR Marc Lehmann http://home.schmorp.de/ =cut 1