Mitel ignoring DNS SRV priority

[jlb5587]

We are using an internet based VOIP provider with a 3300 that we run through a Mitel Border Gateway and they have provided us with a DNS SRV record with a clearly defined priority 1 weight 50 and priority 2 weight 50 for a primary and backup server.

The issue that we are currently experiencing is that the even though the primary is not down it will flip back and forth between the primary and backup at random. At best this may be every hour or so, at worse its several times a minute. So far the only way we can stop this is to just force it directly to an A record but we lose redundancy with this method and isn't optimal.

Has anyone ran into this issue? Is there a setting that we are missing somewhere that is telling the Mitel system to treat these as a round robin?

[Tom2018]

Hi,

where did you entered the weight 50? The weight is for load balancing between two or more MBGs.
In the sip trunk options is a option "DNS SRV resiliency timeout ". There you should enter at wich time the server should be switch, if the first one didn't react
The default is 5 seconds, but may vary from provider to provider

regards,

[jlb5587]

The weight is set in the SIP providers SRV record.

Primary has Priority 1 Weight 50
Backup has Priority 2 Weight 50

Non-authoritative answer:
_sip._udp.ny01-siptrunk-pai.megapathvoice.net   SRV service location:
          priority       = 2
          weight         = 50
          port           = 5060
          svr hostname   = lsancagb-siptrunk-pai.megapathvoice.net
_sip._udp.ny01-siptrunk-pai.megapathvoice.net   SRV service location:
          priority       = 1
          weight         = 50
          port           = 5060
          svr hostname   = nycmny83-siptrunk-pai.megapathvoice.net

lsancagb-siptrunk-pai.megapathvoice.net internet address = 67.103.60.53
nycmny83-siptrunk-pai.megapathvoice.net internet address = 72.244.209.53

DNS SRV REsiliency is set to 30

In this setup I should not be using the backup ever unless primary is lost. But I am still having it flip to the backup. SIP Provider recommended this resiliency setting. Should I try raising the resiliency even higher?

[Tom2018]

Hm,
normally the priority with 1 is the first to take.
the resiliency timer with 30 second is very high; in the worst case an user must wait over 30 second to start a new call.

could you ping the named servers?

[jlb5587]

The servers do not respond to any pings. We figured it would be best because we were having call issues due to the back and forth and was recommended by the sip provider. Will lowering it have a better impact?

[Tom2018]

32 seconds is the highest value you can enter.
Maybe create a wireshark trace and take a look at this, why the server switch over

[x-man]

do a nslookup whilst the problem is having. The DNS server should respond to that.... have two laptops going at once and query both at the same time if you feel the need.

IIRC you can do a nslookup from the command prompt in the maintenance commands as well.