Hello Charles, 

sorry for late reply, we've had some issues rebuilding docker images with kamailo nightly build!

We are running on our local, private on premise kubernetes cluster so connections from outside is simply impossible (in dev we are using a "closed" network system)

I've just tried to : 

- open a tcpdump trace in the instance tracing only the DMQ port defined [5062] - (tcpdump -i eth0 port 5062)

- packet coming from nowhere

- started kamailio ($kamailio -f $PATH_KAMAILIO_CFG -m 512 -DD -E -e)

- seeing this in tcpdump : 

root@router-0:/# tcpdump -i eth0 port 5062

tcpdump: verbose output suppressed, use -v or -vv for full protocol decode

listening on eth0, link-type EN10MB (Ethernet), capture size 262144 bytes

22:06:44.966991 IP router-0.router.paolo.svc.cluster.local.5062 > 42.100.109.113.5062: UDP, length 497

22:06:45.572084 IP router-0.router.paolo.svc.cluster.local.5062 > 42.100.109.113.5062: UDP, length 497

22:06:46.571875 IP router-0.router.paolo.svc.cluster.local.5062 > 42.100.109.113.5062: UDP, length 497

so seems that kamailio is trying to contact the foreign ip

if i check inside the configuration: 

modparam("dmq", "server_address", DMQ_SERVER_ADDRESS)

modparam("dmq", "notification_address", DMQ_NOTIFICATION_ADDRESS)

modparam("dmq", "multi_notify", 1)

modparam("dmq", "num_workers", 4)

modparam("dmq", "ping_interval", 15)

...

  if(is_method("KDMQ") && dst_port == DMQ_PORT){

    dmq_handle_message();

    exit;

  }

and in DMQ_NOTIFICATION_ADDRESS I have : 

#!define DMQ_NOTIFICATION_ADDRESS "sip:dmq-router-service:5062"

dmq-router-service DNS resolves: 

> dmq-router-service.paolo.svc.cluster.local

Server: 10.96.0.10

Address: 10.96.0.10#53

Non-authoritative answer:

Name: dmq-router-service.paolo.svc.cluster.local

Address: 172.28.0.240

so -nowhere- i am able to find 42.100.109.113

and in kamailio log: 

13(821) ERROR: dmq [notification_peer.c:599]: notification_resp_callback_f(): deleting server sip:42.100.109.113:5062 because of failed request

Cheers,

Paolo

2018-07-26 18:14 GMT+02:00 Charles Chance <charles.chance@sipcentric.com>:

Hi Aleksandar,

The initial depopulation of the nodes (following a period of 'pending' state) is due to no response being received from them. Are you able to trace the messages to/from one of them to confirm what is happening there?

As for the unrecognised IP, I'm afraid I can't answer that one. It must be coming from somewhere - it's just a case of finding out where. Following the initial population/depopulation of the node list, DMQ simply repeats the process over and over until it eventually connects to one or more other nodes. Are you running on a public cloud somewhere or is it your own, locally running Kubernetes cluster?

Cheers,

Charles

On 26 July 2018 at 15:30, Aleksandar Sosic <alex.sosic@evosip.cloud> wrote:

Hi Charles,

We're so glad about the improvements you just committed! Thanks!

Now I'm using the latest nightly: 5.2.0~dev6+0~20180726010431.1165+xenial
Kamailio starts even if the DNS record does not exist at first, that's
great. I'm having this nodes up and running:
```
proxy-66f79498cc-8ws6d                3/3       Running   0
4m        172.28.1.4
proxy-66f79498cc-b68dd                3/3       Running   0
4m        172.28.1.5
proxy-66f79498cc-lhbxg                3/3       Running   0
4m        172.28.1.6
proxy-66f79498cc-png6p                3/3       Running   0
4m        172.28.1.7
```

In those 4-5 minutes I've experienced a population and depopulation of
the dmq nodes.

For example:
```
kamcmd> dmq.list_nodes
{
host: 172.28.1.4
port: 5061
resolved_ip: 172.28.1.4
status: pending
last_notification: 0
local: 0
}
{
host: 172.28.1.6
port: 5061
resolved_ip: 172.28.1.6
status: pending
last_notification: 0
local: 0
}
{
host: 40.100.109.113
port: 5061
resolved_ip: 40.100.109.113
status: pending
last_notification: 0
local: 0
}
{
host: 172.28.1.5
port: 5061
resolved_ip: 172.28.1.5
status: active
last_notification: 0
local: 1
}
```

And after a few seconds only the loca node itself:
```
kamcmd> dmq.list_nodes
{
host: 172.28.1.5
port: 5061
resolved_ip: 172.28.1.5
status: active
last_notification: 0
local: 1
}
```
And then again the 3 nodes and the 40.100.109.113 which is a Microsoft
IP and I don't know what's doing that IP into my dmq server list.

The DNS record is this:
```
root@proxy-66f79498cc-b68dd:/# nslookup dmq-proxy-service
Server: 10.96.0.10
Address: 10.96.0.10#53

Non-authoritative answer:
Name: dmq-proxy-service.alex.svc.cluster.local
Address: 172.28.1.7
Name: dmq-proxy-service.alex.svc.cluster.local
Address: 172.28.1.6
Name: dmq-proxy-service.alex.svc.cluster.local
Address: 172.28.1.5
Name: dmq-proxy-service.alex.svc.cluster.local
Address: 172.28.1.4
```
Any clues why that ip is in the dmq nodes list? And why .7 is missing?
Kind regards,
--
Aleksandar Sosic
mail: alex.sosic@evosip.cloud

On Wed, Jul 25, 2018 at 3:06 PM Charles Chance
<charles.chance@sipcentric.com> wrote:
>
> Hi,
>
> I have just pushed some changes to master - one of these allows startup to continue even if initial node resolution fails.
>
> There are some other improvements, too, which I have been planning to push for some time and which should also help in your situation.
>
> Can you try again with these changes applied and let me know the outcome?
>
> Cheers,
>
> Charles

_______________________________________________
Kamailio (SER) - Users Mailing List
sr-users@lists.kamailio.org
https://lists.kamailio.org/cgi-bin/mailman/listinfo/sr-users

--

Charles Chance
Managing Director

t. 0330 120 1200    m. 07932 063 891

Sipcentric Ltd. Company registered in England & Wales no. 7365592. Registered office: Faraday Wharf, Innovation Birmingham Campus, Holt Street, Birmingham Science Park, Birmingham B7 4BB.

_______________________________________________
Kamailio (SER) - Users Mailing List
sr-users@lists.kamailio.org
https://lists.kamailio.org/cgi-bin/mailman/listinfo/sr-users