In article <p45rl61r5m04akh7kh009veqkq15086q57@4ax.com>,
"(PeteCresswell)" <x@y.Invalid> writes:
> The easy part would seem to be:
>
> 1800xxxxxxx<:@gw0>|
>
> But, to be consistent with prepending "1" to 10-digit numbers
> as in:
>
> <1>[2-9]xx[2-9]xxxxxxxS0|
>
> I'd like to combine the two, so that if somebody just
> dials "800-123-4567", the dial plan will prepend a "1"
> and rout it to POTS.
>
>
> Converting literally, I get:
>
> <1>800800xxxxxxx<:@gw0>|
>
> But that looks wrong bc there is nothing to delimit the "look
> for" string and the "convert to" string.
I think you have to do it with two entries:
<800:1800>xxxxxxxS0<:@gw0>|1800xxxxxxxS0<:@gw0>
but I haven't tried these.
Isn't 888 also free in the US?
In that case, you'll need another two entries for 888
--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]
Per (PeteCresswell):
>But that looks wrong bc there is nothing to delimit the "look
>for" string and the "convert to" string.
FWIW, this seems tb the way to do it:
-----------------------------------------------------------
(
911<:@gw0>S0 | Force "911" to POTS, dialing
immediately
1800xxxxxxx<:@gw0>S0 | Immeditate dial to POTS for 1-800
numbers
<:1>800xxxxxxx<:@gw0>S0 | Append "1" to 10-digit 800 calls and
send them out on POTS
1xxx[2-9]xxxxxxS0 | Immediate dial for normal "1 + 10
digits" numbers
<1>[2-9]xx[2-9]xxxxxxxS0 | If number is only 10 digits, prefix
it with "1" and dial immediately
011[2-9]x. Handle overseas dialing
)
-----------------------------------------------------------
But still a question:
How come both of these work, even though
there is a colon missing in the second one?
---------------------------------------------------------------
<:1>800xxxxxxx<:@gw0>S0 | Append "1" to 10-digit 800 cals and
send them out on POTS
<1>[2-9]xx[2-9]xxxxxxxS0 | If number is only 10 digits, prefix
it with "1" and dial immediately
---------------------------------------------------------------
Or am I just wishing that both of them work (I have not figured
out yet how to tell whether the first one is working or just
silently failing and the 800 number is getting processed by
<1>[2-9]xx[2-9]xxxxxxxS0)
--
PeteCresswell
In article <2susl69pihjvpf9vh71g3ll62o8ukd1ra0@4ax.com>,
"(PeteCresswell)" <x@y.Invalid> writes:
> Per Andrew Gabriel:
>> <800:1800>xxxxxxxS0<:@gw0>
>
> Thanks!... after reading your post, the conversion syntax has
> finally dawned on me:
>
> - Conversion enclosed in carrots
> - Value to scan for on left
> - Colon as delim
> - Value to replaced scanned value on right
Yes.
> But, looking at other expressions that seem to work, it seems
> like there is more than one syntax for conversions.
I think it's always as above.
This dialplan language is defined somewhere (I found it once, but
I've forgotten where) and it's got a name, but I can't remember
that either (it's several years since I played with it).
Although it might look quite simple, it's actually very powerful. AFAIK, it wasn't invented by Sipura - they just used it, although
the parts like @gw0 are obviously Sipura additions.
--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]
In article <g3usl61q4ea6cj0u9dsv1brola7i822hs0@4ax.com>,
"(PeteCresswell)" <x@y.Invalid> writes:
> Per (PeteCresswell):
>>But that looks wrong bc there is nothing to delimit the "look
>>for" string and the "convert to" string.
>
> FWIW, this seems tb the way to do it:
> -----------------------------------------------------------
> (
> 911<:@gw0>S0 | Force "911" to POTS, dialing
> immediately
Ideally, you need to test this.
If there's some other short number you can dial on POTS which
is less serious than the police turning up, you could temporarily
swap that in and try it.
> 1800xxxxxxx<:@gw0>S0 | Immeditate dial to POTS for 1-800
> numbers
> <:1>800xxxxxxx<:@gw0>S0 | Append "1" to 10-digit 800 calls and
> send them out on POTS
>
> 1xxx[2-9]xxxxxxS0 | Immediate dial for normal "1 + 10
> digits" numbers
> <1>[2-9]xx[2-9]xxxxxxxS0 | If number is only 10 digits, prefix
> it with "1" and dial immediately
> 011[2-9]x. Handle overseas dialing
> )
> -----------------------------------------------------------
>
> But still a question:
>
> How come both of these work, even though
> there is a colon missing in the second one?
I don't know what <1> means (which doesn't mean it's invalid,
just that I don't know).
> ---------------------------------------------------------------
> <:1>800xxxxxxx<:@gw0>S0 | Append "1" to 10-digit 800 cals and
> send them out on POTS
>
> <1>[2-9]xx[2-9]xxxxxxxS0 | If number is only 10 digits, prefix
> it with "1" and dial immediately
> ---------------------------------------------------------------
>
> Or am I just wishing that both of them work (I have not figured
> out yet how to tell whether the first one is working or just
> silently failing and the 800 number is getting processed by
> <1>[2-9]xx[2-9]xxxxxxxS0)
Temporarily remove the <1>[2-9]xx[2-9]xxxxxxxS0 entry, and see
if it still works.
--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]
"Andrew Gabriel" <andrew@cucumber.demon.co.uk> wrote in message news:ijmgc4$26a$2@news.eternal-september.org...
> In article <g3usl61q4ea6cj0u9dsv1brola7i822hs0@4ax.com>,
> "(PeteCresswell)" <x@y.Invalid> writes:
>> Per (PeteCresswell):
>>>But that looks wrong bc there is nothing to delimit the "look
>>>for" string and the "convert to" string.
>>
>> FWIW, this seems tb the way to do it:
>> -----------------------------------------------------------
>> (
>> 911<:@gw0>S0 | Force "911" to POTS, dialing
>> immediately
>
> Ideally, you need to test this.
> If there's some other short number you can dial on POTS which
> is less serious than the police turning up, you could temporarily
> swap that in and try it.
>
>> 1800xxxxxxx<:@gw0>S0 | Immeditate dial to POTS for 1-800
>> numbers
>> <:1>800xxxxxxx<:@gw0>S0 | Append "1" to 10-digit 800 calls and
>> send them out on POTS
>>
>> 1xxx[2-9]xxxxxxS0 | Immediate dial for normal "1 + 10
>> digits" numbers
>> <1>[2-9]xx[2-9]xxxxxxxS0 | If number is only 10 digits, prefix
>> it with "1" and dial immediately
>> 011[2-9]x. Handle overseas dialing
>> )
>> -----------------------------------------------------------
>>
>> But still a question:
>>
>> How come both of these work, even though
>> there is a colon missing in the second one?
>
> I don't know what <1> means (which doesn't mean it's invalid,
> just that I don't know).
>
>> ---------------------------------------------------------------
>> <:1>800xxxxxxx<:@gw0>S0 | Append "1" to 10-digit 800 cals and
>> send them out on POTS
>>
>> <1>[2-9]xx[2-9]xxxxxxxS0 | If number is only 10 digits, prefix
>> it with "1" and dial immediately
>> ---------------------------------------------------------------
>>
>> Or am I just wishing that both of them work (I have not figured
>> out yet how to tell whether the first one is working or just
>> silently failing and the 800 number is getting processed by
>> <1>[2-9]xx[2-9]xxxxxxxS0)
>
> Temporarily remove the <1>[2-9]xx[2-9]xxxxxxxS0 entry, and see
> if it still works.
>
> --
> Andrew Gabriel
> [email address is not usable -- followup in the newsgroup]
I have a little DTMF digit grabber I can clip across a line to display what is being sent out.
With a 999/112 test, I would use a parallel POT to break dialtone with a suitable digit, to ensure the call fails, not 9 obviously,
and 1 might be undesirable in areas that have access to the NHS 111 service. (I bet that number gets plenty of phantom calls from
the wind
blowing, and shorting dangling pairs, it's well known that 112 suffers from this).
Per Graham.:
>I have a little DTMF digit grabber I can clip across a line to display what is being sent out.
>With a 999/112 test, I would use a parallel POT to break dialtone with a suitable digit, to ensure the call fails, not 9 obviously,
>and 1 might be undesirable in areas that have access to the NHS 111 service. (I bet that number gets plenty of phantom calls from
>the wind
>blowing, and shorting dangling pairs, it's well known that 112 suffers from this).
My 911 test currently consists of changing "911" to "611" - which
connects to a very convenient "We're no longer here and you
should look there...." message.
I think I'm in the market for a "digit grabber".
That plus a little "faux SIP" computer app that I can point the
gateway at instead of my provider - and which displays what it
gets sent....
With those two plus a set of tests to go through each time I
update the dial plan I think I'd feel a lot more comfortable -
not to mention less likely to make so many
"I-don't-have-a-clue-please-help" posts.
--
PeteCresswell
Per (PeteCresswell):
>I think I'm in the market for a "digit grabber".
I seem to have found both the tool in WireShark with a filter.
For the benefit of any other noobs climbing the same curve, here
is what I am doing to test my dial plans:
================================================== ===========
01) Download and install WireShark
02) In the 3102's web interface, select "Admin" and "Advanced"
from the small blue type in the upper right.
03) Set Line 1 | SIP Debug Option: to "Full"
04) Set System | Debug Server to the local IP address of
the PC that you will be running WireShark on.
05) Fire up WireShark
06) (Probably...) observe that packets are now starting to
appear in WireShark's main window.
07) Choose Analyze | Display Filters. A dialog will pop.
08) Make up a "Filter Name"
09) Populate "Filter String" with the expression:
syslog.msg contains "INVITE sip:"
10) Click "Apply"
11) Click "OK"
12) Observe that WireShark's main window has become empty
of packets.
13) Pick up your phone and make a VOIP call. It does not have
to be to a working number.
14) Observe that WireShark's main window, after a brief pause,
is now showing the "INVITE" packet for that call - which
contains the actual number put out by the gateway.
* If the number were to go out to your SIP provider, the
"Info" column would say something like:
INVITE sip:18882345678@callcentric.com....
* If, OTOH, the number were to go out on POTS, the "Info"
column would say something like:
The only joker in this little deck is that WireShark can only
handle so many packets without running out of memory.
My scheme works as far as only showing the desired packets, but
WireShark's total packets keep on incrementing and eventually it
terminates itself gracefully with nice little explanatory
message.
I've got a post in the WireShark forum and I'm hoping to hear
that it's just something dumb I'm doing vis-a-vis the manner of
applying the filter and that there is a way to tell WireShark to
not even save packets that don't meet the filter's criteria.
--
PeteCresswell
Meanwhile, at the uk.telecom.voip Job Justification Hearings,
(PeteCresswell) chose the tried and tested strategy of:
> My scheme works as far as only showing the desired packets, but
> WireShark's total packets keep on incrementing and eventually it
> terminates itself gracefully with nice little explanatory
> message.
What you want is the Capture Filter in Wireshark, which does what you might
think it does.
Capture Options > Capture Filter > "port syslog"
This will only capture port 514/udp.
To be honest I'd never thought of using Wireshark as a syslog server, but it
sounds like it just about works for you, so why not!
--
<http://ale.cx/> (AIM:troffasky) (UnSoEsNpEaTm@ale.cx)
18:33:39 up 3 days, 21:13, 7 users, load average: 0.14, 0.10, 0.08
"I am utterly appalled at how I have been treated like a criminal"
-- Andrew Crossley, ACS:Law, 13 August 2010
Per alexd:
>What you want is the Capture Filter in Wireshark, which does what you might
>think it does.
>
>Capture Options > Capture Filter > "port syslog"
>
>This will only capture port 514/udp.
>
>To be honest I'd never thought of using Wireshark as a syslog server, but it
>sounds like it just about works for you, so why not!
I tried the capture filter thing over-and-over.... but no joy.
I get the distinction between capture and display filters, but
every time I tried, wound up with a display and not a capture
filter.
My workaround has been to use DumpCap.exe - which comes with
WireShark. DumpCap creates a .pcap file, same as WireShark; and
WireShark can read the file.
There seem tb syntax limitations on one type of filter over the
other, and it looks like searching for a substring isn't in the
cards capture-wise.
So I feed DumpCap a capture filter based in the IP of the gateway
and then feed WireShark a display filter that carves down the
gateway's packets.
To Wit:
--------------------------------------------------------------------------
:* ================================================== =====================
:* PURPOSE: To capture network traffic from VOIP adapter and then show
:* same using WireShark to open the capture file.
:* ================================================== =====================
@ECHO OFF
ECHO .
ECHO Ctl-C, then reply "N" to stop and view
ECHO .
C:
CD "C:\Program Files\WireShark
SET DumpLoc=\\NAS\Temp\DumpCap.pcap
dumpcap.exe -w %DumpLoc% -i
\Device\NPF_{35418EFA-22FB-4ADF-A88C-892918610B9F} -f "src net
10.0.0.4"
Meanwhile, at the uk.telecom.voip Job Justification Hearings,
(PeteCresswell) chose the tried and tested strategy of:
> I tried the capture filter thing over-and-over.... but no joy.
>
> I get the distinction between capture and display filters, but
> every time I tried, wound up with a display and not a capture
> filter.
There's either a difference in the way the Windows and Linux versions of
Wireshark implement capture filters, or yours has got a bug in it.
> My workaround has been to use DumpCap.exe - which comes with
> WireShark. DumpCap creates a .pcap file, same as WireShark; and
> WireShark can read the file.
>
> There seem tb syntax limitations on one type of filter over the
> other, and it looks like searching for a substring isn't in the
> cards capture-wise.
Yes, the capture filter language is different to that of the display
filters, which is what I thought you meant by "I get the distinction between
capture and display filters"!
--
<http://ale.cx/> (AIM:troffasky) (UnSoEsNpEaTm@ale.cx)
20:23:58 up 4 days, 23:04, 6 users, load average: 0.00, 0.04, 0.19
"I am utterly appalled at how I have been treated like a criminal"
-- Andrew Crossley, ACS:Law, 13 August 2010
In article <0q06m6lbjk5f83kgr9vm2v8q9d766iqkch@4ax.com>,
"(PeteCresswell)" <x@y.Invalid> writes:
> Per (PeteCresswell):
>>I think I'm in the market for a "digit grabber".
>
> I seem to have found both the tool in WireShark with a filter.
>
> For the benefit of any other noobs climbing the same curve, here
> is what I am doing to test my dial plans:
>
> ================================================== ===========
> 01) Download and install WireShark
>
> 02) In the 3102's web interface, select "Admin" and "Advanced"
> from the small blue type in the upper right.
>
> 03) Set Line 1 | SIP Debug Option: to "Full"
>
> 04) Set System | Debug Server to the local IP address of
> the PC that you will be running WireShark on.
>
> 05) Fire up WireShark
If the system is a unix-like system, you can simply capture the
messages with syslogd (although it can be fiddly as the syslog
header is missing from some of the messages).
If the system is Windows, Spiura used to supply a executable
which captured and logged the messages, although I don't know
if that's still available from Cisco.
--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]
In message <ik43er$oha$1@news.eternal-september.org>, Andrew Gabriel
<andrew@cucumber.demon.co.uk> writes
>
>If the system is Windows, Spiura used to supply a executable
>which captured and logged the messages, although I don't know
>if that's still available from Cisco.
>
If not, I've been using Kiwi's free syslog server for some time: http://www.kiwisyslog.com/kiwi-syslog-server-overview/
--
Nick (=----)
Per alexd:
>There's either a difference in the way the Windows and Linux versions of
>Wireshark implement capture filters, or yours has got a bug in it.
I would not be too quick to rule out user RCI.
--
PeteCresswell
Sending "800" numbers to POTS? 
Linear Mode
