The main configuration to create application groups:
init name {
application_groups [
applicationgroup.name
]
}
E.g.:
init web {
application_groups ["applicationgroup.webstack"]
}
application_group name {
applications [
"application.name"
]
}
E.g.:
application_group webstack {
applications [
"application.db"
"application.www"
]
}
application "name" {
mode application_mode
requires [other applications]
start [executable start args]
stop [executable stop args]
pidfile file
dir working_directory
env {
new var1 value
new var2 value
pass oldname newname
}
env_file env_filename
stdout stream_destination
stderr stream_destination
uid int
gid int
healthchecks [
"healthcheck.name"
]
}
application_mode
can be one of oneshot
, simple
or forking
Applications are started with a clean environment. Environment
variables can be added with env
and env_file
fields.
env
values can take two forms. env new <name> <value>
creates a
new variable with the supplied value. env pass <oldname> <newname>
creates a new variable named newname
with the value taken from
Riffol’s environment variable oldname
.
env {
new {
VAR1 value1
VAR2 value2
}
pass {
ENV1 VAR3
ENV2 VAR4
}
}
This creates two new environment variables, VAR1
and VAR2
with
values value1
and value2
respectively. It also passes the
variables ENV1
and ENV2
into new variables VAR3
nad VAR4
respectively.
env_file <filename>
reads a file of environment variables, one per
line in the following format:
VAR1=value1
VAR2=value2
VAR3
VAR4=
VAR1
and VAR2
are set to value1
and value2
respectively. VAR3
and VAR4
are set to the empty value "".
The env_file
field is processed before the env
field so variables
set up using env
will override those read from env_file
.
stream_destination
can be one of:
file [
filename
]
syslog {
socket unix_sock_address
facility syslog_facility
severity syslog_severity
}
rsyslog {
address remote_inet_address
local local_inet_address
facility syslog_facility
severity syslog_severity
}
syslog_facility
is one of kern
, user
, mail
, daemon
, auth
,
syslog
, lpr
, news
, uucp
, cron
, authpriv
, ftp
, local0
,
local1
, local2
, local3
, local4
, local5
, local6
or
local7
. (default daemon
)
syslog_severity
is one of emerg
, alert
, crit
, err
,
warning
, notice
, info
or debug
(default debug
)
healthcheckfail
can be one of start
, restart
or stop
.
E.g.:
application www {
exec "/etc/init.d/http"
dir "/var/www"
env {
new var1key var1value
new var2key var2value
pass oldname newname
}
env_file "/etc/httpd/morevars"
start start
stop stop
restart restart
stdout file "/var/log/riffol_www.log"
stderr syslog {
}
uid 0
gid 0
healthchecks [
"healthcheck.www"
]
healthcheckfail restart
}
healthcheck name {
checks [
"class://value"
]
interval int
timeout int
}
There are several checks
classes:
-
df
, disk free space -
proc
, process name -
tcp
, TCP connection -
udp
, UDP connection -
http
, establish a http connection -
https
, establish a https connection
Parameters:
-
interval
, the interval of the check defined in seconds -
timeout
, the timeout of network connections defined in seconds
E.g.:
healthcheck db {
checks [
"df:///var/lib/mysql:512"
"proc://mysqld",
"tcp://127.0.0.1:3306"
]
interval 60
timeout 10
}