I am not too fond with CDR but as far as I know the CDR's are sent by the callmanagers, solarwinds would basically need to make a whole new tool to evaluate and use that kind of data to be able to report on it.
unfortunately not as simple as just polling on the data you want. 😞
How would you use the data? We're aware of several different use cases for CDR data; I'm curious as to what yours are and what problems this data would solve for you.
To answer your question, no definite timeframes.
Thanks for replying. I could use a CDR to monitor the call volume coming into my access servers as well as monitor the call durations, the call failures due to hardware issues such as modem failures, repeat call backs, connect speeds, called and calling phone numbers, and reason for disconnect and if possible receive and transmit bytes/bits. A timestamp would be nice too. I need something that can quickly display these call details to identify issues before they become monsters.
I recently had an issue with call durations lasting only 6 seconds when they should have been averaging 40 to 70 seconds on STB call collecting. If I had a CDR, I could have identified this anomaly sooner rather than later.
We also have voice gateways for our VoIP implementation to take calls routed from our circuit providers to our call centers. One issue that comes to mind, where a CDR would be useful is one way audio. When an agent identifies and marks a call as one way audio, we would be able to identify the voice gateway that the troubled call came in through. In this case, we could indentify the hardware or the circuit as possible issues that we can address.
Here are two examples of info from Cisco AS5400XM logs for STB call collecting:
2010-03-12 09:00:19 Local7.Info mem-stb-cc6 11459676: Mar 12 09:00:20.039: %CALLTRKR-6-CALL_RECORD: ct_hndl=3819359, service=PPP, origin=Answer, category=Modem, DS0 slot/port/ds1/chan=7/0/8/20, called=<dddddddddd>, calling=<ssssssssss>, resource slot/port=2/38, userid=xxxxxxx, ip=10.3.173.83, account id=3609790, setup=03/12/2010 08:58:46, conn=0.01, phys=4.98, service=11.35, authen=10.14, init-rx/tx b-rate=2400/2400, rx/tx chars=16846/3874, time=83.54, disc subsys=CSM, disc code=0xE, disc text=Failed to detect carrier
2010-03-12 09:00:16 Local7.Info mem-stb-cc6 11459594: Mar 12 09:00:17.015: %MODEMCALLRECORD-6-PM_TERSE_CALL_RECORD: DS0 slot/contr/chan=7/8/20, slot/port=2/38, call_id=2C2, userid=xxxxxxx, ip=10.3.173.83, calling=<ssssssssss>, called=<dddddddddd>, std=V.22bis, prot=ASYNC, comp=None, init-rx/tx b-rate=2400/2400, finl-rx/tx b-rate=2400/2400, rbs=0, d-pad=None, retr=1, sq=7, snr=0, rx/tx chars=16846/3874, bad=0, rx/tx ec=0/0, bad=0, time=80, finl-state=Steady Retrain, disc(radius)=lost-carrier/No Carrier, disc(modem)=1F06 <unknown>/Requested by host/network indicated disconnect
Our voice gateways that run on similar equipment output information differently, but I’d like similar reports for a CDR.
We are just jumping into the VoIP pool, and CDRs are one of the things that we are looking to provide to our business unit clients as part of this project. I am seriously considering IP SLA Manager anyway for the health monitoring of the network to support VoIP, but CDR would be a huge win if it could provide customizable reports. We aren't looking as much at a charge back model for usage, but the business has expressed a serious interest in gaining insight into the call activity and productivity of the users in their distributed operations.
Of course the basic information would be important - the other thing that's come up in our discovery is the ability to report on multi-leg calls, i.e. if I call the main number for the site, and get the receptionist, and then she transfers me to the operations manager, how can we report on that call as a single customer interaction, even though it landed on multiple phones.
Our company as well is headed more towards a optimal CDR solution that provide good reporting of understandable data. Currently we have Infortel which isnt all that great. In addition to that its expensive and yet another system that we must maintain.
Our company is looking for an overall well designed single platform to get all of our monitoring data and metrics from and today Solarwinds Orion with its plug-ins is looking like a very good solution. However it still is missing some very key elements that I forsee as easy money makers for your company as a whole.
Today from CDR we attempt to extract such things as COSTS and Cost Savings as we are an international company and VoIP internal calls do save the company significant amounts of money. The issue here is that the executives need to be made aware of HOW much is being saved and HOW well its all performing as its their call as to push more money our way to further expand our VoIP system.
Additionally we try (and fail) at being able to create an understandable report that highlights an overall picture of QOS per site and as a whole. This could easily be done with CDR and the RIGHT front end to generating these reports.
If you are seriously considering spawning up new development of a CDR addon for Orion, please add me to the beta team :).
I'l contact you off line about your CDR requirements.
Having others express their CDR interest in this thread will help us prioritizing.
A few word on your use case (accounting, troubleshooting) and the retention needs (should we keep all fields? what retention period?) would be of great help.
Thanks to those who have already written above.
Tks for your input. We are still looking at some form of CDR / CMR support into Orion.
I'd like to get some sense of what use cases are of highest priorities for the community.
If you are interested to "vote" and comment for your favority use case, please do it in this thread, just reply and put your use case ID(s) by decreasing priority order.
E.g. reply "B, A"
if you are primarily interested in CDR for troubleshotting reasons and then accounting reasons and don't care about others.
If you can think of other use cases, also feel free to describe them.
Vote for your favorite CDR Use cases (listed here in no particular order):
-(A) Report on Internal saving /accounting
Sum # of minutes of PSTN calls saved
Needs to understand PSTN rating ($$ for local vs. long distance vs. int'al call)
needs to be able to allocate calls / savings to departments
CDR used to detect user issues (e.g. MOS from CDR)
CMR used to explain bad MOS (jitter? Loss? latency???)
Needs time-based correlation with NPM & NTA stats.
E.g. can interface/PRIs saturation explain bad calls at the same time?
E.g. can CB QOS CoS discarding packets at the same time?
E.g. Can very high Router CPU explain bad calls at the same time?
- (C) Circuit capacity planning
Erlang report (circuit capacity planning). How busy are my b-channels?
- (D) Overall Network Quality characterization
Based on aggregation of all CDR only (detailed CDR not retained)
Characterize the overall network quality, calculate trends
SolarWinds solutions are rooted in our deep connection to our user base in the THWACK® online community. More than 150,000 members are here to solve problems, share technology and best practices, and directly contribute to our product development process. Learn more today by joining now.