Best practices: ping 600 ms

Asterisk Asterisk Support
1

Hello all.
Are there any best practices when the ping between SIP servers is about 600 ms ? They are far away from each other. Some parts of route is organised through the satellite.

All that I founs is ----JITTER BUFFER CONFIGURATION---- section in sip.conf.
As I understand, it’s worth configuring when the ping changes from good to bad. But in our case it’s stable and high:

siprouter*CLI> sip show peers like 7XXXX Name/username Host Dyn Nat ACL Port Status 7XXXXXXXXXXX aa.bb.cc.dd 5062 OK (611 ms)

[root@siprouter /root]# ping -c 5 -S mm.nn.oo.pp aa.bb.cc.dd
PING aa.bb.cc.dd (aa.bb.cc.dd) from mm.nn.oo.pp: 56 data bytes
64 bytes from aa.bb.cc.dd: icmp_seq=0 ttl=56 time=902.888 ms
64 bytes from aa.bb.cc.dd: icmp_seq=1 ttl=56 time=593.485 ms
64 bytes from aa.bb.cc.dd: icmp_seq=2 ttl=56 time=622.970 ms
64 bytes from aa.bb.cc.dd: icmp_seq=3 ttl=56 time=593.661 ms
64 bytes from aa.bb.cc.dd: icmp_seq=4 ttl=56 time=627.913 ms

--- aa.bb.cc.dd ping statistics ---
5 packets transmitted, 5 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 593.485/668.183/902.888/118.225 ms
2

I don’t think there is a best practice on Asterisk directly for deal with high latency on remote peers . You could use narrow-band codec for save some bandwidth, implement QoS or increase your bandwith

3

Generally, the phone itself is responsible for echo cancellation, and it should do that near the source, so that there can be no far end echo. Obviously avoid things like speaker phones, which are difficult to fully echo cancel.