Nortel CDR D records

[RichardLynch]

I am successfully parsing individual CDR records.

I am now "matching up" records for a single "call sequence".

E.g., An N record might be tied to a D record, if they have the same ORIGNO and TERMNO, as a "near" timestamp.

I seem to have a lot of D records in isolation, however.

What exactly would cause a D record anyway?

"Station Activity" is a big vague, to me...

PS
I still don't have MS Excel, so the spreadsheet isn't all that useful...

 

[PERRYPJ]

A D record is produced when CDR is configured for the customer and a trunk
either terminates on, or is accessed by, a set with CDMA class of service
(CLS = CDMA in LD 10 or LD 11).

OLD ROLMEN WORKING ON NORTELS AND AVAYA

 

[RichardLynch]

I get the part that a DMCA device has connected to the trunk...

I'm just not clear on the actual sequence of events that leads to the D record.

For example, there are a lot of N records followed by a corresponding D record. I'm interpreting that to mean that a call came in, and the D device was used to answer it.

Similarly, there are some L records with corresponding D records, I believe.

But then I have a bunch of D records with no corresponding N record. Nor an S and E nor a Q nor anything at all, really.

It's just a lone D record in isolation, as far as I can tell...

So is it possible for there to be a D record that indicates a call was made and answered with no other related records?

Or am I missing some records?

Or am I failing to tie the D records to their correct antecedents?

Here is an example, with real numbers sanitized out:

N 094 00 1205 A001040 01/13 14:16 00:00:22 A 5555555555555
& 0000 0000
D 095 00 1205 A001040 01/13 14:17 00:00:22 555555555555555
N 096 00 A001002 1320 01/13 14:17 00:00:02
& 0000 0000 7034091775XXXXXX
D 097 00 A001002 1320 01/13 14:17 00:00:02
D 098 00 1280 T001046 01/13 14:17 00:00:00 55555
N 099 00 1326 A001039 01/13 14:17 00:00:22 A 55555555555555
& 0000 0000
D 100 00 1212 T001039 01/13 14:17 00:00:02 5555555555

So the first and second record go together, as do the 3rd and 4th.

The 5th and 7th records, are D records I cannot account for.

Is that somehow tied to the others?

Or is it possible for a D record to represent the whole call sequence? My reading of the docs seemed to indicate that the D should have some other record before it to tie it to, but maybe I'm just not getting it...

 

[RichardLynch]

Aha!

Some D records are the whole call sequence, according to the text in the EXAMPLES section of the docs!

Apparently, some routings never generate anything but the D record.

Even more fun, it turns out that there are E (END) records which have no corresponding S (START) records, in direct contradiction to yet another part of the manual.

The E (END) records correspond to D (STATION ACTIVITY) records. Not only that, the E appears in the log *before* the D, which seems pretty backwards to this naive reader, but there it is.

My sample size is only 1340 records, but I can now parse and analyze 100% of them.

Yay!