Quick question, before I read everything: did you designate a "seed" node in your cluster? To achieve this, you must pick a node and set the "flags" column to the "seed" string. We're patching up a lot of documentation during these days, including this "seed" idea.

Liviu Chircu
OpenSIPS Developer

On 04.04.2018 12:11, John Quick wrote:
Hi Liviu,

Here are the modparam settings I'm using:
modparam("usrloc", "working_mode_preset", "full-sharing-cluster")
modparam("usrloc", "location_cluster", 1)
modparam("usrloc", "cluster_mode", "full-sharing")
modparam("usrloc", "restart_persistency", "sync-from-cluster")
modparam("usrloc", "sql_write_mode", "write_back")
modparam("usrloc", "use_domain", 0)
modparam("usrloc", "nat_bflag", "NATTED_CLIENT")
modparam("usrloc", "timer_interval", 61)
modparam("usrloc", "max_contact_delete", 40)

modparam("nathelper", "received_avp", "$avp(rcvip)")
modparam("nathelper", "natping_interval", 38)
modparam("nathelper", "ping_nated_only", 0)
modparam("nathelper", "sipping_bflag", "NAT_SIP_PINGS")
modparam("nathelper", "sipping_from", "sip:pinger@%HOST_IP_ADDR%")

modparam("clusterer", "db_url",
modparam("clusterer", "current_id", %CLUSTER_NODE_ID%)
modparam("clusterer", "node_timeout", 180)

Several things are not working here. I think there must be a fundamental
error in my test setup:
  - The NAT'd UA pinging uses the R-URI of the edge proxy instead of the
Contact address
  - Contact synchronisation does not happen when I re-start one of the
cluster nodes
  - NAT pings are being sent from both nodes in the cluster

I currently have a 2 node cluster (ID's 1 and 3). I don't use "current_info"
and "neighbour_info" to define the cluster nodes because I'm trying to use a
single "master" cfg file which has the smallest possible number of changes
between nodes. Instead, the cluster node data is all stored in the DB table.
On node ID 1:
# opensipsctl fifo clusterer_list
Cluster:: 1
         Node:: 3 DB_ID=2 URL=bin: Enabled=1 Link_state=Up
Next_hop=3 Description=reg3
On node ID 3:
# opensipsctl fifo clusterer_list
Cluster:: 1
         Node:: 1 DB_ID=1 URL=bin: Enabled=1 Link_state=Up
Next_hop=1 Description=reg1

Ideas for what could be wrong:
Is the load order significant for modules such as clusterer, usrloc,
Could my DB schema be wrong? - it was created using v2.4 dev a couple of
weeks ago (not with v2.4 beta as released on 28th March)
Might I not have the latest code - it reports itself as v2.4.0 beta. I
loaded it with the command " git clone
https://github.com/OpenSIPS/opensips.git -b 2.4 opensips-2.4" a few days ago

John Quick
Smartvox Limited
Web: www.smartvox.co.uk

Users mailing list

Reply via email to