Howdy,
I’m running Cisco Call Manager Express across our Mikrotik link. 40-odd phones, a max of 10-odd concurrent sessions.
With Active Directory/Exchange/web browsing running on the background.
VOIP runs on VLANs on Cisco layer3 switches.
VOIP traffic is passed on transparently through the Mikrotik routers.
We have so-far not enabled QOS.
The total link distance is approx 15KM.
4 x Routerboard 433AH - Dual Nstream config in turbomode. 5Ghz.
OfficeA ↔ halfway station ↔ OfficeB
In general call quality is acceptable. 4/5.
But every once in a while call quality drops severely and calls are disconnected.
During a conversation phrases and/or words are re-transmitted or dropped completely.
The Cisco kit marks all VOIP packets with “IP Precendence 5”
My question(s):
How does Cisco’s “IP Precedence 5” translate into Mikrotik language.
Is this the “DSCP/TOS” setting?
If yes, which one?
What do I need to configure in order to have all 4 routerboards prioritize all voice traffic.
Do I need to specify source and destination address ranges or can I set one blanket (cover all) rule for voice traffic?
Any help would be greatly appreciated.
Many thanks.
HendrikF
Cisco and few other sources are saying, that it is ToS setting (ToS is obsolete) and DSCP is being used now.
To translate ToS value to DSCP,
DSCP value=ToS value/4
You can use mangle and mark all traffic with changed DSCP value,
ip firewall mangle add dscp=<used_value> + other required configuration,
http://www.mikrotik.com/testdocs/ros/3.0/qos/mangle.php
Mark VoIP traffic with changed DSCP with one packet-mark,
mark all other traffic with not changed DSCP wit other packet-mark.
Setup 2 queues, one queue for DSCP packet and higher priority=1, second queue for non-dscp traffic and lower priority,
http://www.mikrotik.com/testdocs/ros/3.0/qos/queue.php
Fantastic, thanks guys.
Both replies have helped quite a bit.
VOIP traffic is MUCH better but TCP connections seem to have gone for a dive.
Data flow is MUCH slower now that QOS is running correctly.
cheers
HendrikF
Thanks for the replies guys.
Now that TOS/QOS is working we have been experiencing a lot of random routerboard resets.
This did not happen prior to us enabling TOS.
FI: VOIP handsets at site2 would lose connectivity to the Cisco router at site1 for approx 5 seconds.
Enough to disconnect all calls and force the VOIP handsets to re-negotiate with the router.
Pings sometimes time-out but not always.
It’s almost as-if someone puts a large hand in front of one of the antennas for 5 seconds.
Any ideas on how to resolve the resetting issue would be greatly appreciated.
