beanstalkd
- simple, fast work queue
beanstalkd
[options]
Beanstalkd
is a simple work-queue service. Its interface is
generic, though it was originally designed for reducing the latency of
page views in high-volume web applications by running time-consuming
tasks asynchronously.
When started, beanstalkd
opens a socket (or uses a file descriptor
provided by the init(1) system, see ENVIRONMENT) and listens for
incoming connections. For each connection, it reads a sequence of
commands to create, reserve, delete, and otherwise manipulate "jobs",
units of work to be done. See file doc/protocol.txt
in the
beanstalkd
distribution for a thorough description of the meaning
and format of the beanstalkd
protocol.
-b
path
beanstalkd
will recover any binlog that is present
in path, then, during normal operation, append new jobs and
changes in state to the binlog.-f
ms
A ms value of 0 will cause beanstalkd
to call fsync every time
it writes to the binlog.
The default behavior is to sync every 50 ms.
(This option has no effect without -b
.)
-F
-f
with an infinite ms value.
(This option has no effect without -b
.)
-h
-l
addr
When addr starts with "unix:", the unprefixed value of it will be
used as the local filesystem path to create a UNIX socket instead of
a TCP socket. In this case the value of -p
will be ignored.
(Option -l
has no effect if sd-daemon(5) socket activation is
being used. See also ENVIRONMENT.)
-p
port
(Option -p
has no effect if sd-daemon(5) socket activation is
being used. See also ENVIRONMENT.)
-s
bytes
(This option has no effect without -b
.)
-u
user
-V
-v
-z
bytes
-c
-n
LISTEN_PID
, LISTEN_FDS
sd-daemon(3), sd_listen_fds(3)
Files README.md
and doc/protocol.txt
in the beanstalkd
distribution.
Beanstalkd
is written by Keith Rarick and maintained by the community at
https://github.com/beanstalkd/beanstalkd/issues