VoIP One-way Audio and Voice drops
-
@coliver said:
No, I am looking at it just as a for instance. Have you worked with the ERPoE one? or the Lite one, they look like they have the same hardware minus the additional PoE ports.
Specifically, i have both the ERL and the ERPoE in production.
-
@JaredBusch said:
@coliver said:
No, I am looking at it just as a for instance. Have you worked with the ERPoE one?
They do not have all the same hardware. The additional thing on the PoE is that eth2-4 also have a switch chip. The ERL has no switching chip so if you want two interfaces on the same network for that model, you have to bridge them which in turn reduces max throughput (wakes you down to 300-400 mbps max).
With the PoE you can setup those ports on the switch chip and use them like any other switch. This is how I use it. I have one site with 2 UAP. I have it powering the UAP on the switch ports. eth0 is WAN and eth1 is LAN. sw0 is the WiFi with the eth3 and eth4 being on the switch. eth2 is unused at the moment.
Oh, that is some really good information. I was looking to use it specifically in the fashion you describe and make it my "level 3" switch in addition to the router. Which is how I currently use the Meraki.
-
@coliver said:
Oh, that is some really good information. I was looking to use it specifically in the fashion you describe and make it my "level 3" switch in addition to the router. Which is how I currently use the Meraki.
The ER-Pro does NOT have a switch chip. All ports are routed. It is not designed to act as a switch. Those devices were designed for WISPs in general, though of course anyone can use them.
I have no idea what max throughput would be on a set of bridged interfaces on an ER-Pro 8. It would be higher than an ERL because there is much better processor in it.
-
@coliver said:
@JaredBusch said:
@coliver said:
No, I am looking at it just as a for instance. Have you worked with the ERPoE one?
They do not have all the same hardware. The additional thing on the PoE is that eth2-4 also have a switch chip. The ERL has no switching chip so if you want two interfaces on the same network for that model, you have to bridge them which in turn reduces max throughput (wakes you down to 300-400 mbps max).
With the PoE you can setup those ports on the switch chip and use them like any other switch. This is how I use it. I have one site with 2 UAP. I have it powering the UAP on the switch ports. eth0 is WAN and eth1 is LAN. sw0 is the WiFi with the eth3 and eth4 being on the switch. eth2 is unused at the moment.
Oh, that is some really good information. I was looking to use it specifically in the fashion you describe and make it my "level 3" switch in addition to the router. Which is how I currently use the Meraki.
Best to get a small L3 switch, if that is what you need. Why do you need an L3 switch?
-
@JaredBusch said:
@coliver said:
Oh, that is some really good information. I was looking to use it specifically in the fashion you describe and make it my "level 3" switch in addition to the router. Which is how I currently use the Meraki.
The ER-Pro does NOT have a switch chip. All ports are routed. It is not designed to act as a switch. Those devices were designed for WISPs in general, though of course anyone can use them.
I have no idea what max throughput would be on a set of bridged interfaces on an ER-Pro 8. It would be higher than an ERL because there is much better processor in it.
I'm looking at the datasheets. That info must be buried somewhere else. Thanks for that info I will have to look at the ERPoE then.
-
@scottalanmiller said:
@coliver said:
@JaredBusch said:
@coliver said:
No, I am looking at it just as a for instance. Have you worked with the ERPoE one?
They do not have all the same hardware. The additional thing on the PoE is that eth2-4 also have a switch chip. The ERL has no switching chip so if you want two interfaces on the same network for that model, you have to bridge them which in turn reduces max throughput (wakes you down to 300-400 mbps max).
With the PoE you can setup those ports on the switch chip and use them like any other switch. This is how I use it. I have one site with 2 UAP. I have it powering the UAP on the switch ports. eth0 is WAN and eth1 is LAN. sw0 is the WiFi with the eth3 and eth4 being on the switch. eth2 is unused at the moment.
Oh, that is some really good information. I was looking to use it specifically in the fashion you describe and make it my "level 3" switch in addition to the router. Which is how I currently use the Meraki.
Best to get a small L3 switch, if that is what you need. Why do you need an L3 switch?
I don't think L3 was really what I meant (mostly because that was what I've it described as in the past by a few Cisco techs). I was looking for a switch aggregator or center to the star topology, instead of doing switch daisy chaining.
-
@coliver said:
Oh, that is some really good information. I was looking to use it specifically in the fashion you describe and make it my "level 3" switch in addition to the router. Which is how I currently use the Meraki.
What I do in locations where I have more than one AP but already have PoE injectors is use the ERL and a $20 gigabit switch.
eth0 - WAN - To ISP device
eth1 - LAN - To main LAN switch
eth2 - WiFi - To $20 dumb switchI plug the access points into the dumb switch and I am done. Any dumb switch will blindly pass the VLAN tagging so everything just works.
This is specifically replacing the scenario you described. If the AP gear can be plugged in to the main switch and appropriate tagging and trunking setup on that, there is no need for something like this.
-
@JaredBusch said:
@coliver said:
Oh, that is some really good information. I was looking to use it specifically in the fashion you describe and make it my "level 3" switch in addition to the router. Which is how I currently use the Meraki.
What I do in locations where I have more than one AP but already have PoE injectors is use the ERL and a $20 gigabit switch.
eth0 - WAN - To ISP device
eth1 - LAN - To main LAN switch
eth2 - WiFi - To $20 dumb switchI plug the access points into the dumb switch and I am done. Any dumb switch will blindly pass the VLAN tagging so everything just works.
This is specifically replacing the scenario you described. If the AP gear can be plugged in to the main switch and appropriate tagging and trunking setup on that, there is no need for something like this.
This would replace the Meraki firewall and manage all WAN traffic coming to/from our LAN.
-
In my other thread, this is a similar situation that I am a looking to.
I was/am considering the EdgeRouter-8, though the 5 port would probably do me just fine (assuming no one here has had any issues with VOIP through the 5 port).
-
@Dashrender said:
In my other thread, this is a similar situation that I am a looking to.
I was/am considering the EdgeRouter-8, though the 5 port would probably do me just fine (assuming no one here has had any issues with VOIP through the 5 port).
The only caution on the ER-8 is to be aware that none of the 8 ports are hardware switched.
Because of that, I have never bought one as I never want to give up some throughput by bridging a few interfaces.
The ER PoE is a great choice when you just need a couple switch ports on the router for convenience.
Otherwise, the router is really not the place to want switched ports in the first place.
-
Replaced the firewall. Still seeing the same issues we were before.
-
@coliver said:
Replaced the firewall. Still seeing the same issues we were before.
This needs qualified.
Replaced how? Swapped a Meraki unit? that woudl imply same programming thus potentially the same issue. Completely different hardware? Then it comes to verifying the new configuration.
-
@JaredBusch said:
@coliver said:
Replaced the firewall. Still seeing the same issues we were before.
This needs qualified.
Replaced how? Swapped a Meraki unit? that woudl imply same programming thus potentially the same issue. Completely different hardware? Then it comes to verifying the new configuration.
Completely new firewall - ERPoE-5. I'm running into the same issues I was before with latency and packet loss, symptoms are exactly the same.
-
At this point you are really pointing to the ISP.
Let's think here.
You swapped router.
You swapped SIP trunk provider.
You swapped from PBX to direct on a phone.Potential solutions to try:
Have your ruled out the local switching hardware.
Have you ruled out needing QoS on the LAN? Obviously this is extremely rare, but you have already tested every normal source of an issue.
Can you connect from a secondary ISP at all on site? -
@JaredBusch said:
At this point you are really pointing to the ISP.
Let's think here.
You swapped router.
You swapped SIP trunk provider.
You swapped from PBX to direct on a phone.Potential solutions to try:
Have your ruled out the local switching hardware.Wired the PBX (which is a VM) directly to the router, via a different port on the server and a new Hyper-V virtual switch dedicated to just the PBX virtual machine. Still encountered the same issues. This was prior to the recent router switch. I'm considering bringing up a second host to test it out on.
Have you ruled out needing QoS on the LAN? Obviously this is extremely rare, but you have already tested every normal source of an issue.
It seems to only affect calls to and from the outside world. Would local QoS provide
Can you connect from a secondary ISP at all on site?
No, unfortunately we are very rural which makes a different ISP impossible, we only have one option for a SIP trunk provider for our numbers... which is the ISP.
-
QoS is not very likely as the issue is not quality, but dropping.
-
Are you sure that STUN is configured?
-
@scottalanmiller said:
Are you sure that STUN is configured?
I am fairly certain STUN isn't configured, nor do I know how to go about doing that. With STUN don't both end points (our SIP trunk and PBX) have to be configured with the same STUN server?
-
@scottalanmiller said:
Are you sure that STUN is configured?
Why do you bring up STUN again? this has nothing to do with STUN. The phones are internal to the PBX.
-
@coliver said:
@scottalanmiller said:
Are you sure that STUN is configured?
I am fairly certain STUN isn't configured, nor do I know how to go about doing that. With STUN don't both end points (our SIP trunk and PBX) have to be configured with the same STUN server?
Wait, when STUN is a necessity, why are we going through all this troubleshooting if the basics aren't done yet. I said earlier that if STUN wasn't set up this would happen.
