Management server installation
Since version 1.6.3-175 we started to use central configuration server to store yeti module configuration for all nodes in cluster.
Packages installation
# apt update && apt install yeti-management
Configuration files
/etc/yeti/management.cfg
This file contains configuration for management daemon.
Set desired logging level and address to listen.
You can set multiple addresses separated by comma to listen on multiple interfaces.
Possible log_level values are: (1 - error, 2 - info, 3 - debug)
daemon {
listen = {
"tcp://0.0.0.0:4444"
}
sctp {
address = "127.0.0.1"
port = 4444
}
http {
address = "127.0.0.1"
port = 3000
}
log_level = 2
}
/etc/yeti/system.cfg
This file contains configuration for all nodes. Each top-level section defines configuration for node of certain type (signaling is for traffic switch nodes). All top-level sections contains mandatory section globals which must have all possible values common for all nodes. Then there is named sections for each node_id which may contain values overriding ones set in global section.
Note: even if your node does not have any specific values you have to define empty section for this node anyway, otherwise management node will not provide configuration for node with this id.
Example of minimal configuration file for node with id 8
signalling {
globals {
yeti {
pop_id = 4
msg_logger_dir = /var/spool/sems/dump
log_dir = /var/spool/sems/logdump
routing {
schema = switch17
function = route_release
init = init
master_pool {
host = 127.0.0.1
port = 5432
name = yeti
user = yeti
pass = yeti
size = 4
check_interval = 10
max_exceptions = 0
statement_timeout=3000
}
failover_to_slave = false
slave_pool {
host = 127.0.0.1
port = 5432
name = yeti
user = yeti
pass = yeti
size = 4
check_interval = 10
max_exceptions = 0
statement_timeout=3000
}
cache {
enabled = false
check_interval = 60
buckets = 100000
}
}
cdr {
dir = /var/spool/sems/cdrs
completed_dir = /var/spool/sems/cdrs/completed
pool_size = 2
schema = switch
function = writecdr
master {
host = 127.0.0.1
port = 5432
name = cdr
user = cdr
pass = cdr
}
failover_to_slave = false
slave {
host = 127.0.0.1
port = 5432
name = cdr
user = cdr
pass = cdr
}
failover_requeue = true
failover_to_file = false
serialize_dynamic_fields = true
batch_size = 1
batch_timeout = 5000
}
resources {
reject_on_error = false
write {
host = 127.0.0.1
port = 6379
size = 2
timeout = 500
}
read {
host = 127.0.0.1
port = 6379
size = 2
timeout = 1000
}
}
registrations {
check_interval = 5000
}
rpc {
calls_show_limit = 10000
}
}
}
node 8 { }
}
Warning
You should define all Nodes that has been created at web interface by their IDs. See System->Nodes menu. SEMS node will refuse to start if corresponding node is not defined at system.cfg
Management server launch
Launch configured management server instance:
# service yeti-management start
Checks
Check file /var/log/yeti/yeti-management.log for daemon logs:
# tail -2 /var/log/yeti/yeti-management.log
Sep 12 12:54:47 evial yeti-management[25376]: [25376]
info: server/src/yeti_mgmt_server.cpp:148: starting version 1.0.5
Sep 12 12:54:47 evial yeti-management[25376]: [25376]
info: server/src/mgmt_server.cpp:123: listen on tcp://0.0.0.0:4444
Check listening port:
# netstat -lpn | grep yeti_management
4444 tcp 0 0 0.0.0.0:4444 0.0.0.0:* LISTEN 25376/yeti_manageme