I have this issue when from time to time all of the peers become unreachable. We first thought this was because of our UDP buffer was to small since we saw spikes in UDP InErrors when that happened, but after increasing that it fixed the UDP InErrors but not the unreachable peers. Most of the answers I found on this forum where to disable “Qualify” but I’m not sure that’s the best way, while testing that out it doesn’t go directly to voicemail when a peer is really offline.
We also notice that it happens with our Asterisk servers that have the “heaviest” load. They are located in different data centers so it doesn’t seem like it’s a problem with our server provider. The servers are not behind NAT and are dedicated servers. The phones are also all in different locations. There are around 500 peers on the servers we are encountering problems.
All the symptoms indicate your network is overloaded, so you really need to address that. You might be able to get some symptomatic relief if you ncrease the value of qualify, but that will only work if packets are being delayed, not dropped.
If you control the whole network, you can set quality of servfice parameters and configure the routers to priorities traffic marked with them. This will not help if the congestion is on part of the network you don’t control, as the public internet generally doesn’t honour thse parameters, as they would be likely to be absued.
So best thing to do is to contact the data center to fix their network? The data centers we use are from the same company but in different locations so it could be a common problem with their setup.
If the network is controlled by your company, you should ask them if they have implemented, or can implement, differentiated services oan the network, and which options are enabled.
They may want you to provide information on your traffice needs for both SIP signalling and media.