Discussion:
[cisco-voip] Calculating MOS using CAR / CDR / SQL
Robert Kulagowski
2007-09-11 21:03:28 UTC
Permalink
CM 4.1.3sr2

I'm trying to calculate MOS for one of my sites (VM calls to Unity over
the WAN)

When I looked in CAR, it appears that I'll only get "Good" "Acceptable",
etc, which isn't what I'm looking for - I really need to have some number.

Using the data that's in the SQL database, I should be able to do a SQL
query to pull the calls from the site that go do VM, but how do I
correlate that to the records that details jitter, etc? I can't seem to
find the thread that ties the CallDetailRecord table with the one that
actually stores the raw data.
Wes Sisk
2007-09-12 17:39:48 UTC
Permalink
yes, this is a pseudo classification that CAR applies based on jitter,
packet loss, etc. It examines the stats for the call to reach this value.

later versions of CM get a true MOS calculation from the phones in the
CDR/CAR database in table calldetailrecorddiagnostics in columns
origvarVQMetrics and destvarVQMetrics

sample value:
MLQK=4.5000;MLQKav=4.5000;MLQKmn=4.5000;MLQKmx=4.5000;ICR=0.0000;CCR=0.0000;ICRmx=0.0000;CS=0;SCS=0;MLQKvr=0.95

for 4.x use the CDR definitions doc to correlate rows beteween
calldetailrecord and calldetailrecorddiagnostic:
http://www.cisco.com/en/US/customer/products/sw/voicesw/ps556/products_programming_usage_guide09186a00806556fb.html

/Wes
Post by Robert Kulagowski
CM 4.1.3sr2
I'm trying to calculate MOS for one of my sites (VM calls to Unity over
the WAN)
When I looked in CAR, it appears that I'll only get "Good" "Acceptable",
etc, which isn't what I'm looking for - I really need to have some number.
Using the data that's in the SQL database, I should be able to do a SQL
query to pull the calls from the site that go do VM, but how do I
correlate that to the records that details jitter, etc? I can't seem to
find the thread that ties the CallDetailRecord table with the one that
actually stores the raw data.
_______________________________________________
cisco-voip mailing list
https://puck.nether.net/mailman/listinfo/cisco-voip
Robert Kulagowski
2007-09-12 20:24:11 UTC
Permalink
Post by Wes Sisk
yes, this is a pseudo classification that CAR applies based on jitter,
packet loss, etc. It examines the stats for the call to reach this value.
later versions of CM get a true MOS calculation from the phones in the
CDR/CAR database in table calldetailrecorddiagnostics in columns
origvarVQMetrics and destvarVQMetrics
MLQK=4.5000;MLQKav=4.5000;MLQKmn=4.5000;MLQKmx=4.5000;ICR=0.0000;CCR=0.0000;ICRmx=0.0000;CS=0;SCS=0;MLQKvr=0.95
for 4.x use the CDR definitions doc to correlate rows beteween
http://www.cisco.com/en/US/customer/products/sw/voicesw/ps556/products_programming_usage_guide09186a00806556fb.html
Hrmm, I'm getting bogus information that's making this harder. When I
did a select * from CallDetailRecordDiagnotic where latency > 0

I only received 573 rows. I would have thought that _every_ call would
would have > 1 latency, so this seems low.

A WAN call should have latencies in the 60-100ms range for these sites.
Wes Sisk
2007-09-12 21:20:32 UTC
Permalink
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Because phones do not implement RTCP latency calculations are not
possible. If you look at all latency values you will find most to be 0.<br>
<br>
/Wes<br>
<br>
Robert Kulagowski wrote:
<blockquote cite="mid:***@gmail.com" type="cite">
<pre wrap="">Wes Sisk wrote:
</pre>
<blockquote type="cite">
<pre wrap="">yes, this is a pseudo classification that CAR applies based on jitter,
packet loss, etc. It examines the stats for the call to reach this value.

later versions of CM get a true MOS calculation from the phones in the
CDR/CAR database in table calldetailrecorddiagnostics in columns
origvarVQMetrics and destvarVQMetrics

sample value:
MLQK=4.5000;MLQKav=4.5000;MLQKmn=4.5000;MLQKmx=4.5000;ICR=0.0000;CCR=0.0000;ICRmx=0.0000;CS=0;SCS=0;MLQKvr=0.95


for 4.x use the CDR definitions doc to correlate rows beteween
calldetailrecord and calldetailrecorddiagnostic:
<a class="moz-txt-link-freetext" href="http://www.cisco.com/en/US/customer/products/sw/voicesw/ps556/products_programming_usage_guide09186a00806556fb.html">http://www.cisco.com/en/US/customer/products/sw/voicesw/ps556/products_programming_usage_guide09186a00806556fb.html</a>
</pre>
</blockquote>
<pre wrap=""><!---->
Hrmm, I'm getting bogus information that's making this harder. When I
did a select * from CallDetailRecordDiagnotic where latency &gt; 0

I only received 573 rows. I would have thought that _every_ call would
would have &gt; 1 latency, so this seems low.

A WAN call should have latencies in the 60-100ms range for these sites.
_______________________________________________
cisco-voip mailing list
<a class="moz-txt-link-abbreviated" href="mailto:cisco-***@puck.nether.net">cisco-***@puck.nether.net</a>
<a class="moz-txt-link-freetext" href="https://puck.nether.net/mailman/listinfo/cisco-voip">https://puck.nether.net/mailman/listinfo/cisco-voip</a>
</pre>
</blockquote>
</body>
</html>

Loading...