Webclient performance Issues

[commercial1]

We have a new deployment that is hosted on OVH the customer has 30mb leased line just for 3cx with 30 extensions. The customer uses the web client latest version of chrome and a desk phone Fanvil X4. They are reporting issues that if they try and answer a call using the webclient it fails to pickup on the desk phone all the time. Sometimes it works fine !

I have tested internal calls between extensions and have noted that at times when you click call on an internal extension it states trying the dialling box then disappears off screen and comes back 1-3 minutes later stating ringing and the call initiates. I have tried creating local DNS entries for the hosted system but this has not helped, the PBX hosted in OVH has been restarted, chrome has been updated, removed and reloaded.

Any one come across this before ?

 

[JohnS_3CX]

Hi @commercial1

Please share some more information:

• How are the phones connected to the hosted instance? (STUN/SBC/VPN)
• What is the PBX version
• What is the Fanvil X4 firmware version
• Are they sure they are on CSTA mode or Web Client mode? (when problem occurs)

 

[commercial1]

They are connected STUN
3cx V16.0.910
Fanvil X4 2.10.0.6646
CSTA mode

Sometimes it in instant but other times can take upto 2 mins to call an internal extension, it shows trying then disappears and pops back up with dialling

 

[Evolute IT]

Did you use different SIP/RTP ports for each phone? If you have 30 extensions with a phone each, use an SBC to fix most issues.

 

[commercial1]

Yes each handset is set for different ports for SIP and RDP so are unique for each handset, they are not heavy users of the system and normally max out at about 4sc calls at any one point in time, it does not appear to be worse or better the amount of calls in use. The memory and processor on the PBX are all low usage. We are trying to avoid deploying a SBC into the system as they are after a true cloud system with no reliance on internal hardware that could fail, if we have no other option then we may go down this route.

 

[Evolute IT]

Yes each handset is set for different ports for SIP and RDP so are unique for each handset, they are not heavy users of the system and normally max out at about 4sc calls at any one point in time, it does not appear to be worse or better the amount of calls in use. The memory and processor on the PBX are all low usage. We are trying to avoid deploying a SBC into the system as they are after a true cloud system with no reliance on internal hardware that could fail, if we have no other option then we may go down this route.

How many phones are at that site?

 

[commercial1]

How many phones are at that site?

Just counted up and 23

 

[Evolute IT]

Just counted up and 23

Then I strongly suggest you install an SBC. Those are very reliable and for your setup you can use the Raspberry version.

This would prevent having to manage all port forwarding and it might fix your CSTA issue as the SBC will proxy it.

 

[commercial1]

Ok i i will get one configured and see if it helps, i was hoping to do it without but calls seem fine its just the delay on the web client that is anoying them.

 

[Evolute IT]

Ok i i will get one configured and see if it helps, i was hoping to do it without but calls seem fine its just the delay on the web client that is anoying them.

The delay is probably caused by so many remote phones using ports at the same time. CSTA requests get lost.

I understand your feeling of adding hardware but it's not worse than a router with a VPN. The router can still fail anytime.

 

[commercial1]

The delay is probably caused by so many remote phones using ports at the same time. CSTA requests get lost.

I understand your feeling of adding hardware but it's not worse than a router with a VPN. The router can still fail anytime.

Yes i was hoping there wouldn't be a delay as there is a 100mb leased line used just for requests to the hosted PBX. Yes for a small office with low budget i agree with single point of failure im fortunate this site has 2 100mb leased lines with cluster firewall and a fttc backup line as well hence the reason for trying to avoid installing a sbc, i might just install one and leave one on the shelf there just incase its ever needed

 

[Evolute IT]

Yes i was hoping there wouldn't be a delay as there is a 100mb leased line used just for requests to the hosted PBX. Yes for a small office with low budget i agree with single point of failure im fortunate this site has 2 100mb leased lines with cluster firewall and a fttc backup line as well hence the reason for trying to avoid installing a sbc, i might just install one and leave one on the shelf there just incase its ever needed

From what I've seen on the forum, those who use Raspberry's never had issues with them. The failure rate is pretty freaking low.

Also, the SBC can be monitored from the MC and you'll get an email if it is down for more than 5min so you have a heads up before the customer calls.

 

[JohnS_3CX]

You should also take a look at your firmware revision, it's not the latest supported for your PBX.

There is a number of improvements that were made since the previous ROM version, including STUN compatibility with 3CX

2.10.0.6646 -> 2.10.1.6836

 

[commercial1]

You should also take a look at your firmware revision, it's not the latest supported for your PBX.

There is a number of improvements that were made since the previous ROM version, including STUN compatibility with 3CX

2.10.0.6646 -> 2.10.1.6836

Thanks we did update the firmware and it didn't seem to help at all, we have also had reports since the firmware upgrade of one way audio and quiet calls, even though we have all the port redirection set for each handset. As a result we are deploying a SBC this week so fingers crossed this just fixes everything

 

[JohnS_3CX]

Yes the SBC should help especially with the audio issues. Let us know how it goes