NAME
Beekeeper::Service::LogTail - Buffer log entries
VERSION
Version 0.08
SYNOPSIS
my $entries = Beekeeper::Service::LogTail->tail(
count => 100,
level => LOG_DEBUG,
host => '.*',
pool => '.*',
service => 'myapp-foo',
message => 'Use of uninitialized value',
after => now() - 10,
);
print "$_->{message}\n" foreach @$entries;
Beekeeper::Service::LogTail->tail_async(
count => 100,
on_success => sub {
my ($entries) = @_;
print "$_->{message}\n" foreach @$entries;
},
on_error => sub {
my ($error) = @_;
die $error->message;
},
);DESCRIPTION
By default all workers use a Beekeeper::Logger logger which logs errors and warnings both to files and to a topic log/{level}/{service} on the message bus.
LogTail workers keep an in-memory buffer of every log entry sent to these topics in every broker of a logical message bus. Then this buffer can be queried using the tail method provided by this module or using the command line client bkpr-log.
By default the buffer holds the last 20000 log entries (this can be changed setting the buffer_entries configuration option of workers). Buffered entries consume 1.5 kiB for messages of 100 bytes, increasing to 2 KiB for messages of 500 bytes. Holding the last million log entries in memory will consume around 2 GiB (!).
LogTail workers are CPU bound and can collect up to 20000 log entries per second. Applications exceeding that traffic will need another strategy to consolidate log entries from brokers.
LogTail workers are not created automatically. In order to add a LogTail worker to a pool it must be declared into config file pool.config.json.
METHODS
tail ( %filters )
Returns all buffered entries that match the filter criteria.
The following parameters are accepted:
count: Number of entries to return, default is last 10.
level: Minimal severity level of entries to return.
host: Regex that applies to worker host.
pool: Regex that applies to worker pool.
service: Regex that applies to service name.
message: Regex that applies to error messages.
after: Return only entries generated after given timestamp.
tail_async ( %filters, on_success => $cb, on_error => $cb )
Asynchronous version of tail method.
Callbacks on_success and on_error must be coderefs and will receive respectively Beekeeper::JSONRPC::Response and Beekeeper::JSONRPC::Error objects as arguments.
SEE ALSO
bkpr-log, Beekeeper::Service::LogTail::Worker.
AUTHOR
José Micó, jose.mico@gmail.com
COPYRIGHT AND LICENSE
Copyright 2015-2021 José Micó.
This is free software; you can redistribute it and/or modify it under the same terms as the Perl 5 programming language itself.
This software is distributed in the hope that it will be useful, but it is provided “as is” and without any express or implied warranties. For details, see the full text of the license in the file LICENSE.