Phone unreachable after a transfer

[Cedric]

Hi everybody,

We have a problem with our phones :

- When a call is received from outside, an operator answers the call and attempts to transfer the call to an internal number. Sometimes, this transfer does not work, the recipient phone is unreachable. After that the phone does not work for 5 or 10 minutes (impossible to call this phone from anywhere). Or, if the unreachable phone makes a new call, the phone operate again just after. I precise everything was well before, this is a very strange issue and we cannot reproduce it all times.

Details if I am not clear :

0. Everything is OK.
1. 014844241XX is calling my company with the number 014844241YY.
2. 4 internal numbers are ringing (200 to 204) and the number 202 answers.
3. The 202 transfers this call to the internal number 205, but this is impossible because 205 seems unreachable (nothing happens).
4a. After that, the number 205 is unreachable between 5 or 10 minutes.
4b. After that, the number 205 is unreachable but if the number 205 tries to make a new call, it will work again.

We had a look to the network and we did not find any relevant information (errors, drops, etc), we tried in a busy hour and in a quiet hour : this is the same. Everything else with the network seems OK.

With the Wireshark capture, we see some ICMP "242 Destination unreachable (port unreachable)" and TCP Retransmission or DUP ACK, I am looking for something with that.


Do you have any ideas about that ?


Phone :
Platform : 6865i
Firmware Version   4.2.0.2023
Firmware Release Code   SIP
Boot Version   Boot2 1.0.0.D Jun 16 2016 01:28
Server MITEL MiVoice Office 400


Thank you and have a good day !

Cédric

[ralph]

Is it possible to temporarily turn off ICMP in your switches and routers and retest?
I'm thinking it may be related.

Ralph

[Cedric]

Hi Ralph and thank you for your answer !

I will try this, but do you think this could have an impact on other things for the MITEL server ?
You think this because it is a possible bug or something like that ?

I will let you know the result, thank you so much !

Cédric

[ralph]

It shouldn't unless you're using ICMP to control your network routing.

Ralph

[Cedric]

Hello,

In the future that would be a problem because I will implement IP SLA and supervision, but now it is ok.
In our opinion, what could be the problem with ICMP ? If the problem is finally ICMP, do you think that turned off ICMP would be the only solution ?

ICMP is turned off since the last Friday, for the moment two persons have faced a problem, but different and I am not sure if it is related.

I will do a more accurate test this evening after working hours, I will let you know the result.

Thank you !

[Cedric]

Hello,

We tried yesterday and unfortunately we was not able to reproduce the transfer problem with the ICMP and without the ICMP.

However, we noticed when we have the ICMP active : 12 "ICMP destination unreachable (port unreachable)", between the both phone, then 1 or 2 "ICMP destination unreachable (port unreachable)" between the phone recipient and the Server (EIP slot 1-1).
The ICMP seems not work in any case for this transfer, but the transfer was working this time (we tried more than 10 times)

I don't know if this is an interesting information, I am still investigating.

Thanks

[Cedric]

Hi,

We have found something interesting I think :

Our server is configured with :
Main address : 192.168.17.10
EIP slot 1-1 : 192.168.17.11

We have reproduced a problem and we have seen some packets received or answered with 17.10 or 17.11, this looks random. We got some malformed or collision packets in the same time, so probably something is wrong with that, but at the moment I don't know why...

Still on it, thanks !

[Cedric]

Hi,

After the intervention of an external company on the problem, we did not found the cause. Below their description if someone has an idea ?

"At about 11:00 one of the phones stopped responding to web browsing. This phone belonged to XX who is based behind the switch 06. The issue with the phone is that whilst it responds to pings the device itself becomes unresponsive. We are not able to logon to the phone interface, pull down captures, change settings. At the same time, the phone becomes unstable when trying to make calls, change the settings, retrieve call lists, etc. Its almost as if the phone freezes.

After a few moments the phones becomes normal again. This time can vary and also, any previous changes or operations on the phone then activate. For example, if a user tries to call someone a few time whilst frozen, the phone will try and call that person. This prevents the phone from being used, whilst also locking out the recipient's phone until the call is cleared. The only way to clear the issues is to reboot the caller's phone and wait for the issues to leave the recipients phone.

We then concentrated on 3 users who were all then starting to have issues. These users are based in the recipient office and the production office (again, not on the same network switch as the phone system). After an hour were we are able to produce some captures of the phones and the issues they were having."


"This was a call from on of the users on the network. The IP address ending .11 is the EIP card which handles call resources."


Thank you for your help,

Cédric

[bojo387]

We have a "slightly" similar case with Mitel SIP terminals. Some 6865i phones freeze or hang. They can't be called or make calls but everything appears normal on the phone. However, the partner could still log in to the phones via web UI. Restart of the phones does not fix the issue and for some reason, the only way was to restart the Mitel 470 every time.

Mitel asked our partner to change "Session refresh timer for active line supervision (s)" from 3600 to 90 secs. or 12 secs., restart the CS, and restart all SIP phones. (Configuration > IP Network > VoIP > SIP)  The case is still being monitored if it reoccurs.

There was another case but with "Standard SIP" terminals. After doing a *# feature command, the terminal could not be called until it makes a call first. For this, the fix was to disable the "Terminal supports session replacement" under Connection Settings of the Standard SIP terminal.

Just throwing these 2 cases in here in case they help.

BR/Bojo

[Cedric]

Hi,

Thanks a lot, I will try that and I will let you know.

Cédric

[Cedric]

Hi,


I tried this settings but that was the same for us. After that, we tried the last update for this kind of Mitel phone and that was ok for something like 2 weeks, we thought we found the solution.
However, we got now the same kind problem again, it is like if the phones are unregistered or lost the connection for a while (5/10min). Sometimes that will be fine and the day after the problem will appear several times, this behaviour are only for the phone, everything else in the network seems to be ok.


The phones are :
Mitel 6865i
Firmware    4.3.0.2036
They got IP addresses from a DHCP server with a mask and a gateway, nothing else.
The VLAN is directly configured on the phone.
The network port are configured with native VLAN data, tagged VLAN voice. Sometimes we have a computer behind the phones but not all times.
The Mitel Server is a MiVoice Office 400.
Our network is Cisco with the core network Cisco 4500X VSS.
The interface voice VLAN is configured with an IP address, dhcp helper and no icmp redirects (we tried without this command this is the same).


If somebody have another idea...

Thank you !