Posted: Tue Apr 15, 2003 4:03 pm Post subject: [Asterisk-Dev] single bit T1 errors (Zhone again)
This is sort of a followup to my ALIT testing problems.
I have a problem where sometimes a call rings through to extensions, and
after the first ring, gets hung up on. I was watching this in the console
just now, and what happened was asterisk thought one of the other phones
was answered and immediately hung up (not the phone beside me but one in a
group that ring together.) Now, no one else was even here at the time. What
I am thinking actually happened was due to the ringing or whatever else
going on with the extension, an offhook condition was incorrectly detected
for a split second, but long enough to answer and then disconnect the call.
I think the solution is essentially the same as for the ALIT testing (no,
not getting it disabled like I ended up doing), but being able to ignore a
couple of bad bits that represent less than a specified interval of time.
I would be more than happy to work on this, but as I said in my last email,
I need some help with what order the events in the driver are expected, and
how to establish timing on them.
Does anyone else with the zhones see this sort of stuff ?
Incidentally this seemed to get a lot worse when I upgraded the asterisk
box to a computer that is much faster.
Posted: Tue Apr 15, 2003 7:54 pm Post subject: [Asterisk-Dev] single bit T1 errors (Zhone again)
Quote:
I think the solution is essentially the same as for the ALIT testing (no,
not getting it disabled like I ended up doing), but being able to ignore a
couple of bad bits that represent less than a specified interval of time.
I would be more than happy to work on this, but as I said in my last email,
I need some help with what order the events in the driver are expected, and
how to establish timing on them.
zaptel.c would be the place for this. I believe there is some code there
already designed to do some minimal debouncing you might take a look at.
Posted: Tue Apr 15, 2003 8:28 pm Post subject: [Asterisk-Dev] single bit T1 errors (Zhone again)
ok fair enough, I have looked but I am not clear on several things
Mark - would you be good enough to answer :
1) does every signalling bit that comes along generate an event in the
driver, or only the first bit of a given signalling type (ringing,
hookstate, etc.) ? If so great, if not what handles this part ?
2) do the events in 1) come in at a reliable interval as defined by the
media properties eg: T1 would have a certain fixed event rate per channel,
and E1 would have another predictable but possibly different rate ?
At 02:54 PM 4/15/2003 -0500, you wrote:
Quote:
> I think the solution is essentially the same as for the ALIT testing (no,
> not getting it disabled like I ended up doing), but being able to ignore a
> couple of bad bits that represent less than a specified interval of time.
>
> I would be more than happy to work on this, but as I said in my last email,
> I need some help with what order the events in the driver are expected, and
> how to establish timing on them.
zaptel.c would be the place for this. I believe there is some code there
already designed to do some minimal debouncing you might take a look at.
Posted: Tue Apr 15, 2003 9:50 pm Post subject: [Asterisk-Dev] single bit T1 errors (Zhone again)
On Tuesday, April 15, 2003 3:28 PM, Jon Pounder [SMTP:JonP@inline.net]
wrote:
Quote:
1) does every signaling bit that comes along generate an event in the
driver, or only the first bit of a given signaling type (ringing,
hookstate, etc.) ? If so great, if not what handles this part ?
I don't know if this is useful information as it is not driver specific
but is T1 specific. Framing and channel configuration have an effect on
signaling states. A signaling state is determined by looking at a
complete superframe, which will contain 2 or 4 signaling bits.
If framing on a T1 is
ESF -
4 signaling bits A, B, C and D (16 possible state) -
superframe is 24 frames long. Signaling come from frames 6,12,18,24
SF -
2 signaling bits A and B - (4 possible states)
superframe is 12 frames long. Signaling bits come from frames 6 and
12.
Quote:
2) do the events in 1) come in at a reliable interval as defined by
the
Quote:
media properties eg: T1 would have a certain fixed event rate per
channel, and E1 would have another predictable but possibly different
rate ?
There would be a reliable interval for the state of a T1, one
superframe. One frame is 1/8000 of a second (125 micro second). So
every 1/8000 second there could be a new state.
I have no idea how the driver passes on signaling. It could be bits or
states or change in states. Also, I do not know if there is any
performance monitoring that could be used to determine if bits are
being received in error.
If any of this is useful great, if not sorry to have used the
bandwidth.
Posted: Tue Apr 15, 2003 10:12 pm Post subject: [Asterisk-Dev] single bit T1 errors (Zhone again)
Thanks Don,
You saved me a bit of math - I understand the basics of how the signalling
works, but as you, I am also unsure of whether the events in the driver
represent every bit that comes along, or just the changes in state of the bits.
I know there are a bunch of other people on the list with zhones, am I the
only one plagued with these issues, or has everyone else just tossed theirs
in the garbage or what ?
At 04:50 PM 4/15/2003 -0500, you wrote:
Quote:
On Tuesday, April 15, 2003 3:28 PM, Jon Pounder [SMTP:JonP@inline.net]
wrote:
> 1) does every signaling bit that comes along generate an event in the
> driver, or only the first bit of a given signaling type (ringing,
> hookstate, etc.) ? If so great, if not what handles this part ?
I don't know if this is useful information as it is not driver specific
but is T1 specific. Framing and channel configuration have an effect on
signaling states. A signaling state is determined by looking at a
complete superframe, which will contain 2 or 4 signaling bits.
If framing on a T1 is
ESF -
4 signaling bits A, B, C and D (16 possible state) -
superframe is 24 frames long. Signaling come from frames 6,12,18,24
SF -
2 signaling bits A and B - (4 possible states)
superframe is 12 frames long. Signaling bits come from frames 6 and
12.
> 2) do the events in 1) come in at a reliable interval as defined by
the
> media properties eg: T1 would have a certain fixed event rate per
> channel, and E1 would have another predictable but possibly different
> rate ?
There would be a reliable interval for the state of a T1, one
superframe. One frame is 1/8000 of a second (125 micro second). So
every 1/8000 second there could be a new state.
I have no idea how the driver passes on signaling. It could be bits or
states or change in states. Also, I do not know if there is any
performance monitoring that could be used to determine if bits are
being received in error.
If any of this is useful great, if not sorry to have used the
bandwidth.
But note, that it appears to be a powerpoint presentation that has been
distilled to pdf form. It says confidential on the print out, but I
found it via google, search "q931 pri" second page, second entry.
On Tue, 2003-04-15 at 17:12, Jon Pounder wrote:
Quote:
Thanks Don,
You saved me a bit of math - I understand the basics of how the signalling
works, but as you, I am also unsure of whether the events in the driver
represent every bit that comes along, or just the changes in state of the bits.
I know there are a bunch of other people on the list with zhones, am I the
only one plagued with these issues, or has everyone else just tossed theirs
in the garbage or what ?
At 04:50 PM 4/15/2003 -0500, you wrote:
>On Tuesday, April 15, 2003 3:28 PM, Jon Pounder [SMTP:JonP@inline.net]
>wrote:
> > 1) does every signaling bit that comes along generate an event in the
> > driver, or only the first bit of a given signaling type (ringing,
> > hookstate, etc.) ? If so great, if not what handles this part ?
>
>I don't know if this is useful information as it is not driver specific
>but is T1 specific. Framing and channel configuration have an effect on
>signaling states. A signaling state is determined by looking at a
>complete superframe, which will contain 2 or 4 signaling bits.
>
>If framing on a T1 is
>ESF -
> 4 signaling bits A, B, C and D (16 possible state) -
> superframe is 24 frames long. Signaling come from frames 6,12,18,24
>SF -
> 2 signaling bits A and B - (4 possible states)
> superframe is 12 frames long. Signaling bits come from frames 6 and
>12.
>
>
> > 2) do the events in 1) come in at a reliable interval as defined by
>the
> > media properties eg: T1 would have a certain fixed event rate per
> > channel, and E1 would have another predictable but possibly different
> > rate ?
>
>There would be a reliable interval for the state of a T1, one
>superframe. One frame is 1/8000 of a second (125 micro second). So
>every 1/8000 second there could be a new state.
>
>I have no idea how the driver passes on signaling. It could be bits or
>states or change in states. Also, I do not know if there is any
>performance monitoring that could be used to determine if bits are
>being received in error.
>
>If any of this is useful great, if not sorry to have used the
>bandwidth.
>
>Don Pobanz
>_______________________________________________
>Asterisk-Dev mailing list
>Asterisk-Dev@lists.digium.com
>http://lists.digium.com/mailman/listinfo/asterisk-dev
Posted: Tue Apr 15, 2003 10:46 pm Post subject: [Asterisk-Dev] single bit T1 errors (Zhone again)
I'll be "demoting" my Zhone to the basement of my house to use as a home
system soon. I'll keep you posted on any similar problems.
Quote:
I know there are a bunch of other people on the list with zhones, am I the
only one plagued with these issues, or has everyone else just tossed
theirs
in the garbage or what ?
Posted: Wed Apr 16, 2003 2:51 am Post subject: [Asterisk-Dev] single bit T1 errors (Zhone again)
Quote:
1) does every signalling bit that comes along generate an event in the
driver, or only the first bit of a given signalling type (ringing,
hookstate, etc.) ? If so great, if not what handles this part ?
It's the transitions, sometimes debounced, which generate events.
Quote:
2) do the events in 1) come in at a reliable interval as defined by the
media properties eg: T1 would have a certain fixed event rate per channel,
and E1 would have another predictable but possibly different rate ?
They're measured about the same frequency. Interpretations often are time
dependent (i.e. FLASH vs. ON HOOK for exmaple).
Posted: Wed Apr 16, 2003 1:57 pm Post subject: [Asterisk-Dev] single bit T1 errors (Zhone again)
List:
One of our two Zhones is less than perfect. Several of the fxs ports do
not respond and we have not been able to isolate why _inbound_ calls are
cut off after 50 seconds. The latter is a problem when customers call but
handy when tele-marketers call :)
Howard White
Quote:
I'll be "demoting" my Zhone to the basement of my house to use as a home
system soon. I'll keep you posted on any similar problems.
> I know there are a bunch of other people on the list with zhones, am I
> the only one plagued with these issues, or has everyone else just
> tossed theirs
> in the garbage or what ?
One of our two Zhones is less than perfect. Several of the fxs ports do
not respond and we have not been able to isolate why _inbound_ calls are
cut off after 50 seconds. The latter is a problem when customers call but
handy when tele-marketers call :)
Just a note here. You are connecting them with T100P cards. I just had
to send one of my T100Ps back to Digium due to it having a lot of line
errors. If your bad setup is the one you have in TN, I may be able to
help you diagnose it by visiting you and bringing my T100P card that is
working flawlessly with me. I only suggest this because I so not trust
the Zhone to do decent diagnostic. I could also bring my ADIT600 with me
also to do good diagnostics.
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum