VoIP One-way Audio and Voice drops
-
@Dashrender said:
It will be interesting to see how it continues to show as the end of the day approaches and passes.
If yesterday or Monday was any indication, it generally works fine. At least compared to the rest of the day.
-
@coliver said:
@Dashrender said:
It will be interesting to see how it continues to show as the end of the day approaches and passes.
If yesterday or Monday was any indication, it generally works fine. At least compared to the rest of the day.
I am more curious about the packet loss as the day winds down... does it drop to zero or near zero, or does it stay high?
-
@Dashrender said:
@coliver said:
@Dashrender said:
It will be interesting to see how it continues to show as the end of the day approaches and passes.
If yesterday or Monday was any indication, it generally works fine. At least compared to the rest of the day.
I am more curious about the packet loss as the day winds down... does it drop to zero or near zero, or does it stay high?
I'll let you know. Right now I just did 170 packets and saw 34% packet loss.
-
-
Definitely sounds like something is overloaded.
-
Yep, @GregoryHall is helping me narrow it down. Just waiting on my ISP to get back to me so we can setup something he requested. Small town ISPs aren't the most responsive I've come to understand.
-
@coliver said:
Yep, @GregoryHall is helping me narrow it down. Just waiting on my ISP to get back to me so we can setup something he requested. Small town ISPs aren't the most responsive I've come to understand.
Depends on which one. Some of them are the best ones out their.
-
-
Hyper-V seems to have some issues reporting Linux memory (although from conversations with @scottalanmiller it seems most things do).
On the Linux Box (PBX)
Hyper-V Manager
-
Ping results from the PBX to Google from late last night until I got in this morning.
-
@coliver said:
Ping results from the PBX to Google from late last night until I got in this morning.
Wow, that is bad.
Did you have a ping running from another device to google at the same time for comparison?
-
@JaredBusch said:
@coliver said:
Ping results from the PBX to Google from late last night until I got in this morning.
Wow, that is bad.
Did you have a ping running from another device to google at the same time for comparison?
No, not at that point unfortunately.
-
@coliver said:
Hyper-V seems to have some issues reporting Linux memory (although from conversations with @scottalanmiller it seems most things do).
On the Linux Box (PBX)
Hyper-V Manager
Yes, most platforms don't read the data in the expected way.
You have tons of free memory here.
-
Ok, after @GregoryHall I was able to setup a third party SIP trunk and that has been working phenomenally.
However with our primary SIP trunk, after some reconfiguration (or something) on the providers side I am now getting this:
This happens only on incoming calls from our primary trunk. Is this a configuration on the router or from the provider that would effect this?
-
Nevermind that was my own mistake. Forgot to configure the Asterisk SIP settings to the new IP address. Odd that it didn't affect both trunks though.
-
Ok, so after changing the IP address to the current one I am now getting a list of these on the primary trunk (again works fine for the secondary). This is accompanied with terrible garbled/robotic audio.
Oddly those ports are way above the range for RTP that I have setup in the PBX server.
-
I almost hate to resurrect an ancient topic, but all of these issues have been resolved, and I would be amiss to not do a post-mortem on the whole resolution.
I think some of the problems with call quality went away, but there was still a lot of trouble with outgoing calls failing (mostly in the afternoon). What we came to realize was that Vitelity (SIP provider) required, but never stated, that it was necessary to have separate trunks set up for inbound and outbound calls. While their site provided the inbound PEER info, there was no posted settings (at least, that any of us could find) dictating the outbound call trunk configuration.
What was happening was that any outbound packets were being de-prioritized to guarantee inbound traffic. From the logs, you just saw that the call went out and got refused by their SIP server. We ended up working with support to learn that we needed an outbound trunk. After setting up the trunk (guessing; they gave us no details) and having the problem persist, we were finally able to give them the trunk settings, which they looked at and said "oh, this isn't correct" and gave us the right configuration. Suddenly, all of our troubles vanished.
I was unaware that certain SIP providers require a trunk for inbound and also outbound traffic.
There was also a secondary issue, which may have played into the latency that was experienced. This has also since been resolved. We had a bottleneck inside the LAN, as a group of switches for all of the users was connected to the "servers" switch, where the PBX lives, with a single 1gbit connection, as well as the gateway connection. Old switches (un-stackable) and physical location had a bit to do with the layout. Once we upgraded to a stack of Netgear S3300's and included the gateway, servers, and users in the "stack", the latency disappeared.
-
That's very good Vitelity information to have. Would be crappy to face that same problem again due to a lack of documentation.
-
I have never heard of a provider that required a separate trunk for inbound and outbound. that is just crazy.
-
And unexpected. How do they expect you to know that you need that?
