Contents
- Using Command-line Options
- Creating a Configuration File
- Setting Server Options
- Setting Resources Options
- Setting Logging Options
- Setting Tuning Options
- Setting Cluster Options
- Setting Internal Options
You can set these options from the command-line; they affect only
the specific instance of the running server:
--port 5672 Server port for clients
--queue_timeout 0 Timeout for auto-deleted queues
--max_memory_mb 512 Maximum allowed memory, MBytes
--per_client 0 Default messages per client
--monitor 0 Monitor interval, seconds
--dump_state 60 Dump state interval, seconds
--record_stats 0 Record performance statistics
--log_path logs Active log file directory
--keep_logs 1 Keep log files
--archive_path archive Archive log file directory
--archive_cmd value Archive log file command
--alert_log alert.log Error log file name
--daily_log daily.log Daily log file name
--debug_log debug.log Debug log file name
--debug_route 0 Debug message routing?
--debug_queue 0 Debug queue activity?
--debug_peering 0 Debug peering messages?
--debug_console 0 Debug console I/O?
--trace 0 Protocol trace level
--heartbeat 2 Heartbeat timer, seconds
--polling_threads 4 Number of polling OS threads
--working_threads 4 Number of working OS threads
--tcp_nodelay 1 TCP/IP NODELAY option
--tcp_rcvbuf 0 TCP/IP receive buffer, in bytes
--tcp_sndbuf 0 TCP/IP send buffer, in bytes
--frame_max 32768 Maximum size of AMQP content frames
--cluster_vhost / Cluster virtual host
--is_primary 0 Server is HA primary
--is_backup 0 Server is HA backup
--primary_peer HA primary peer address
--backup_peer HA backup peer address
--auto_crash 0 Auto-crash test timer, seconds
--auto_block 0 Auto-freeze test timer, seconds
--kernel_limit 0 SMT kernel limit
You can also set the above options in an XML text file, to affect
all future instances of the running server. To create a new XML
configuration file, copy the following block to an editor session,
and then:
-
Delete all options you want to leave as default.
-
Delete this comment and update your name and the date.
-
Save file as new amq_server.cfg or merge with existing copy.
-
Restart the server.
<?xml version="1.0"?>
<!--
Revised: yyyy/mm/dd by yourname
-->
<config>
<server
port = "5672"
queue_timeout = "0"
/>
<resources
max_memory_mb = "512"
per_client = "0"
monitor = "0"
dump_state = "60"
record_stats = "0"
/>
<logging
log_path = "logs"
keep_logs = "1"
archive_path = "archive"
archive_cmd = "value"
alert_log = "alert.log"
daily_log = "daily.log"
debug_log = "debug.log"
debug_route = "0"
debug_queue = "0"
debug_peering = "0"
debug_console = "0"
trace = "0"
/>
<tuning
heartbeat = "2"
polling_threads = "4"
working_threads = "4"
tcp_nodelay = "1"
tcp_rcvbuf = "0"
tcp_sndbuf = "0"
frame_max = "32768"
/>
<cluster
cluster_vhost = "/"
is_primary = "0"
is_backup = "0"
primary_peer = ""
backup_peer = ""
/>
<internal
auto_crash = "0"
auto_block = "0"
kernel_limit = "0"
/>
</config>
-
port = "5672"
-
Specifies the port on which the server should open its connections.
Default value is 5672.
-
queue_timeout = "0"
-
Specifies the delay, in seconds, after which auto-deleted queues are
effectively deleted, following disconnection of their last consumer.
If zero, queues are deleted immediately. Default value is 0. This
option can be changed at runtime.
-
max_memory_mb = "512"
-
Specifies the maximum memory allowed for the server, in MB. When the
server reaches this limit, it will slow down publishers. If zero, no
limit is enforced. Default value is 512. This option can be changed at
runtime.
-
per_client = "0"
-
Specifies the limit per client, in messages. Clients that exceed this
limit will be disconnected, and their messages discarded. Applies only
to messages held on private queues belonging to one client. Default
value is 0. This option can be changed at runtime.
-
monitor = "0"
-
Specifies the interval in seconds at which the server will report its
message rates. If zero, no monitoring is shown. The recommended value
is 1, if monitoring is desired. Default value is 0. This option can be
changed at runtime.
-
dump_state = "60"
-
Specifies the interval at which the server will report its state. This
shows the number of messages, queues, consumers, etc. used by the
server. If zero, no state is logged. Default value is 60. This option
can be changed at runtime.
-
record_stats = "0"
-
Specifies whether or not the server records performance statistics in
a format suitable for importing into a spreadsheet. If set, the server
creates a file called 'amq_server_stats_(portnumber).del' which
contains performance statistics in a comma-delimited format. For this
setting to take effect, the monitor option must also be set. Default
value is 0. This option can be changed at runtime.
-
log_path = "logs"
-
Specifies the directory path into which active log files are placed.
Default value is logs.
-
keep_logs = "1"
-
Specifies whether or not the server should keep logs. If zero, the
server will not keep logs. If 1, it will keep an alert log, a daily
log, and a debug log as specified. Default value is 1.
-
archive_path = "archive"
-
Specifies the directory path into which archived log files are placed.
This can be the same as the active path; archived log files are
renamed to avoid conflict with active log files. Default value is
archive.
-
archive_cmd = "value"
-
If specified, this command is run on all archived log files. You can
use this option to compress archived log files, or clean up old and
unwanted files. When the command is run, the name of the archived log
file is provided to it as a single argument. Default value is not
defined.
-
alert_log = "alert.log"
-
The name of the alert log file, which records all errors and warnings
issued by the server. Note that the server port number is always added
to the name of the log file. Default value is alert.log.
-
daily_log = "daily.log"
-
The name of the daily log file, which records all normal activity as
well as errors. Note that the server port number is always added to
the name of the log file. Default value is daily.log.
-
debug_log = "debug.log"
-
The name of the debug log file, which records all normal activity,
errors, and debug trace information. Note that the server port number
is always added to the name of the log file. Default value is
debug.log.
-
debug_route = "0"
-
Specifies whether exchange routing is logged or not. Set this option
when you are debugging a message routing design. For production use,
we recommend you do not set this option. Default value is 0. This
option can be changed at runtime.
-
debug_queue = "0"
-
Specifies whether queue dispatching is logged or not. Set this option
when you are debugging message processing in the server. For
production use, we recommend you do not set this option. Default value
is 0. This option can be changed at runtime.
-
debug_peering = "0"
-
Specifies whether peering activity is logged or not. Set this option
when you are configuring or debugging peerings. For production use, we
recommend you do not set this option. Default value is 0. This option
can be changed at runtime.
-
debug_console = "0"
-
Specifies whether console activity is logged or not. Set this option
when you are configuring or debugging the console. For production use,
we recommend you do not set this option. Default value is 0. This
option can be changed at runtime.
-
trace = "0"
-
Specifies the protocol debug level, which can be 0 (no tracing), 1
(show information and protocol methods), 2 (show protocol animation),
3 (show wire-level data). You should not use any tracing on production
servers except in low-volume scenarios. Default value is 0. This
option can be changed at runtime.
-
heartbeat = "2"
-
Defines the timeout for connection heartbeating. Default value is 2.
This option can be changed at runtime.
-
polling_threads = "4"
-
On multithreaded builds, defines the number of OS threads dedicated to
socket polling. Default value is 4.
-
working_threads = "4"
-
On multithreaded builds, defines the number of OS threads dedicated to
processing, that is, tasks other than socket polling. Default value is
4.
-
tcp_nodelay = "1"
-
If this value is 1, socket data is written immediately, which is
usually good for latency. If this value is 0, data is buffered until
there is a full packet, which is usually good for throughput. Default
value is 1. This option can be changed at runtime.
-
tcp_rcvbuf = "0"
-
If this value is greater than zero, all client connections will use
the specified value. Note: setting this value is delicate, do not use
this option unless you know what you are doing. Default value is 0.
This option can be changed at runtime.
-
tcp_sndbuf = "0"
-
If this value is greater than zero, all client connections will use
the specified value. Note: setting this value is delicate, do not use
this option unless you know what you are doing. Default value is 0.
This option can be changed at runtime.
-
frame_max = "32768"
-
Defines the maximum frame size the server will accept. Larger frame
sizes are useful for applications that create very large message
headers but will cause congestion on multi-channel connections.
Default value is 32768. Minimum value is 2048. Maximum value is
4194304.
-
cluster_vhost = "/"
-
Specifies the cluster vhost name, which must be identical for both HA
peers. Any string, up to 255 chars long, may be used. By convention
virtual host names are formatted as POSIX paths, starting with '/'.
Note that all applications must use the same virtual host name when
connecting to the cluster. Default value is /.
-
is_primary = "0"
-
Specifies whether or not this server instance is the HA primary peer.
If this option is set, the cluster backup-peer option must be set as
well. Default value is 0.
-
is_backup = "0"
-
Specifies whether or not this server instance is the HA backup peer.
If this option is set, the cluster primary-peer option must be set as
well. Default value is 0.
-
primary_peer = ""
-
Specifies the primary high-availability peer address. This should
consist of a servername:portnumber combination that allows the backup
peer to correctly address the primary peer. Default value is .
-
backup_peer = ""
-
Specifies the backup high-availability peer address. This should
consist of a servername:portnumber combination that allows the primary
peer to correctly address the backup peer. Default value is .
-
auto_crash = "0"
-
If set, the server will crash at irregular intervals, being some
random point within the number of seconds specified. This is used
specifically to test applications for failover and robustness. Set to
zero for all real usage. Default value is 0. This option can be
changed at runtime.
-
auto_block = "0"
-
If set, the server will block at irregular intervals, being some
random point within the number of seconds specified. This is used
specifically to test applications for failover and robustness. Set to
zero for all real usage. Default value is 0. This option can be
changed at runtime.
-
kernel_limit = "0"
-
Sets an action limit for the SMT kernel: after this many steps it will
exit. Use this to break the server for failure testing. Set to zero
for all real usage. Default value is 0. This option can be changed at
runtime.
