ISDN trouble: PRI got event: HDLC Abort (6) on Primary D-channel of span 1

[mikeboss]

this is with ASKOZIA 2.1.3 running on Deciso ISDN (ALIX.2D with OpenVox B400M)
connected to SWISSCOM ISDN-NT1+2ab

i'm getting this error every +/- 30 seconds (NOT exactly 30 secs!)

PRI got event: HDLC Abort (6) on Primary D-channel of span 1

at some point the ISDN goes down with red alert. after two days of trial and
error i was fed up with fiddling around and went down to ASKOZIA 1.0.3 which
seems to work without any issues. i tried all kinds of stuff mentioned at the
openvox wiki, nothing seems to solve the problem.

http://wiki.openvox.cn/index.php/Troubleshooting_of_PRI_cards


regards,
michael

[Sebastian]

Hi,

can you send me your chan_dahdi.conf? Could you also increase your debug and verbose levels (AMI -> core set debug 10, core set verbose 10) and include the logs in your mail?

Thanks 

[mikeboss]

hi sebastian!

since the customer had to open his new office today and i installed Askozia 1.0.3 there's no possibility to get logfiles anymore.
maybe i'll go back there with the Deciso (with he customer's config in it and Askozia 2.1.3 installed) when i have some spare time...

all i can provide at this point is the chan_dahdi.conf:

; Hardware Info
; isdn te port, basechannel 1 (Port 1) is on card OpenVox B400P w/3 channels
; isdn te port, basechannel 4 (Port 2) is on card OpenVox B400P w/3 channels
; isdn te port, basechannel 7 (Port 3) is on card OpenVox B400P w/3 channels
; isdn te port, basechannel 10 (Port 4) is on card OpenVox B400P w/3 channels

[trunkgroups]

[channels]
; Provider: SWISSCOM ISDN Port 1, Port Port 1
signalling=bri_cpe_ptmp
rxgain=0
txgain=0
echocancel=no
language=de-de
context=ISDN-PROVIDER-11722747214e5b8907b46cf-incoming
group=1
channel => 1-2


regards,
michael

[user469]

Hi Sebastian,

as you know i've got some problem with isdn and dahdi , but  i'm also having the problem of mikeboss :
Sep  6 09:31:08 asterisk[1122]: NOTICE[1144]: chan_dahdi.c:2983 in my_handle_dchan_exception: PRI got event: HDLC Abort (6) on D-channel of span 1
Sep  6 09:31:18 asterisk[1122]: NOTICE[1144]: chan_dahdi.c:2983 in my_handle_dchan_exception: PRI got event: HDLC Abort (6) on D-channel of span 1
Sep  6 09:31:29 asterisk[1122]: NOTICE[1144]: chan_dahdi.c:2983 in my_handle_dchan_exception: PRI got event: HDLC Abort (6) on D-channel of span 1
Sep  6 09:31:39 asterisk[1122]: NOTICE[1144]: chan_dahdi.c:2983 in my_handle_dchan_exception: PRI got event: HDLC Abort (6) on D-channel of span 1

I've just updated last night to 2.1.3 as suggested by you in the dahdi post .

Hereunder my chan_dahdi.conf
; Hardware Info
; isdn te port, basechannel 1 (Port 1) is on card OpenVox B400P w/3 channels
; isdn te port, basechannel 4 (Port 2) is on card OpenVox B400P w/3 channels
; isdn te port, basechannel 7 (Port 3) is on card OpenVox B400P w/3 channels
; isdn nt port, basechannel 10 (Port 4) is on card OpenVox B400P w/3 channels

[trunkgroups]

[channels]
; Provider: Isdn1, Port Port 1
signalling=bri_cpe_ptmp
rxgain=0
txgain=0
echocancel=64
language=it-it
context=ISDN-PROVIDER-1232703144cf374d643702-incoming
group=1
internationalprefix=00
nationalprefix=0
localprefix=02
channel => 1-2

; Phone: 101
callerid=Default Extension <101>
signalling=bri_net_ptmp
transfer=yes
callwaiting=yes
callwaitingcallerid=yes
facilityenable=yes
overlapdial=yes
rxgain=0
txgain=0
echocancel=128
language=it-it
context=ISDN-PHONE-1972329334cf371755ba14
group=10
channel => 10-11

Regards

[Sebastian]

Are you using Deciso hardware too?
I am starting to think that it might be a hardware related problem.

@michael: I didn't forget you. Please give me a little more time.

[user469]

ASKOZIA 2.1.3 running on Deciso ISDN (ALIX.2D with OpenVox B400M)
connected to Italy TELECOM ISDN-NT+2ab

Its the Deciso , the OpenVox , the Telco provider or the NT+2ab ??

[giovanni.v]

I am starting to think that it might be a hardware related problem.

Not sure it is an hardware problem, dahdi still has some serius problems handling properly ptmp isdn lines; looking in the digium issue tracker everyone can easily find a lot of (still open) acknowleged tickets related to isdn framing in ptmp setups.

However, an hdlc abort is not necessarily a symptom of failure... if it happens occasionally (http://www.erg.abdn.ac.uk/~gorry/eg3567/dl-pages/hdlc.html).

Hdlc aborts that happens at regular intervals may be caused by the driver not properly handling the idle line state; this is part of euroisn standard and many telcos use it to save power.
Also timing errors may cause framing errors then hdlc aborts.

The first thing I would suggest to users living in europe is to set switchtype=euroisdn in the isdn provider advanced settings.

To test if this is caused by timing errors check dmesg to see if the dahdi drivers switch the clock to another port because the actual sync source goes idle:
[ ... ] wcb4xxp 0000:00:0c.0: new card sync source: port 1
[ ... ] ...
[ ... ] wcb4xxp 0000:00:0c.0: new card sync source: port 2
but the new one may be also idle... because all ports are idle.

Finally check if hdlc aborts happens only when there are not active calls, after a defined time (the idle timeout is 15 seconds in Italy). This should confirm that is a failure caused by the idle line state.

The last resort will be a pri debug ;-)

[Sebastian]

The first thing I would suggest to users living in europe is to set switchtype=euroisdn in the isdn provider advanced settings.

That might be a good idea. On the other hand, it has worked for both of them before. If the switchtype was wrong, it should have never worked.

I just found this:

https://issues.asterisk.org/jira/browse/DAHLIN-239

It has been an open issue for quite some time - never a good sign. 

[mikeboss]

hi sebastian et all

no, at least for me, it never worked (Deciso with Askozia 2.1.3 and ISDN). but it seems to be running flawless now (Deciso with Askozia 1.0.3 installed) which points to a software related problem.

IIRC i tried "switchtype=euroisdn" and it didn't help.


regards,
michael

[giovanni.v]

The first thing I would suggest to users living in europe is to set switchtype=euroisdn in the isdn provider advanced settings.

That might be a good idea. On the other hand, it has worked for both of them before. If the switchtype was wrong, it should have never worked.

Things never are so simple  the default is able to work because belongs to a minimal standard subset but for the same reason may fail on many features:
1) no proper handling of some switch specific protocol features.
2) no idle line state handling (as i know it's an etsi/euroisdn only specification).
Regarding the latter one some differences also exists in the way different low level drivers handle this (e.g. wcb4x versus zaphfc)
3) moon revolutions??? 

https://issues.asterisk.org/jira/browse/DAHLIN-239

It has been an open issue for quite some time - never a good sign. 

Thera are similar issues from 2009, still opened/acknowledged/unresolved.

[user469]

Me too ,  i tried "switchtype=euroisdn" and it didn't help .
Nothing found on dmesg to see if the dahdi drivers switch the clock .

But apart from the Abort messages the pbx is working except sometime
http://forums.askozia.com/index.php/topic,1674.0.html

[pierrot74]

same issue on Decisio :
dahdi_pri_error: 1 Unable to receive TEI from network in state 2(Assign awaiting TEI)!
ISDN provider: France Orange
port 1: no incoming call received
trying to set TEI=0
trying switchtype=euroisdn  also VN3
nothing to do !
alarms RED,
port 3 OK

dmesg:
dahdi: Version: 2.4.1.2
dahdi_transcode: Loaded.
dahdi_echocan_oslec: Registered echo canceler 'OSLEC'
wcb4xxp 0000:00:0c.0: probe called for b4xx...
wcb4xxp 0000:00:0c.0: Identified OpenVox B400P (controller rev 1) at 00011c00, IRQ 9
IRQ 9/b4xxp: IRQF_DISABLED is not guaranteed on shared IRQs
wcb4xxp 0000:00:0c.0: NOTE: hardware echo cancellation has been disabled
wcb4xxp 0000:00:0c.0: Port 1: TE mode
wcb4xxp 0000:00:0c.0: Port 2: TE mode
wcb4xxp 0000:00:0c.0: Port 3: TE mode
wcb4xxp 0000:00:0c.0: Port 4: NT mode
wcb4xxp 0000:00:0c.0: Did not do the highestorder stuff
vzaphfc: HFC-S PCI A ISDN (V1.42) loading
DAHDI Dynamic Span support LOADED
wcb4xxp 0000:00:0c.0: new card sync source: port 4
wcb4xxp 0000:00:0c.0: new card sync source: port 3
wcb4xxp 0000:00:0c.0: new card sync source: port 2
wcb4xxp 0000:00:0c.0: new card sync source: port 1
wcb4xxp 0000:00:0c.0: new card sync source: port 2
wcb4xxp 0000:00:0c.0: new card sync source: port 1
wcb4xxp 0000:00:0c.0: new card sync source: port 4
wcb4xxp 0000:00:0c.0: new card sync source: port 3
wcb4xxp 0000:00:0c.0: new card sync source: port 2
All TDMoE multiframe span groups are active.
eth0: link up, 100Mbps, full-duplex, lpa 0xCDE1
dahdi: Registered tone zone 2 (France)
wcb4xxp 0000:00:0c.0: new card sync source: port 2
__ratelimit: 1 callbacks suppressed
wcb4xxp 0000:00:0c.0: new card sync source: port 2
wcb4xxp 0000:00:0c.0: new card sync source: port 4
wcb4xxp 0000:00:0c.0: new card sync source: port 2

-----------------------------------------------------------------
~ # cat /etc/asterisk/chan_dahdi.conf
; Hardware Info
; isdn te port, basechannel 1 (Port 1 T0 FT) is on card OpenVox B400P w/3 channels
; isdn te port, basechannel 4 (Port 2) is on card OpenVox B400P w/3 channels
; isdn te port, basechannel 7 (Port 3) is on card OpenVox B400P w/3 channels
; isdn nt port, basechannel 10 (Port 4 vers PABX) is on card OpenVox B400P w/3 channels

[trunkgroups]

[channels]
; Provider: FT Orange, Port Port 1 T0 FT
signalling=bri_cpe_ptmp
rxgain=0
txgain=0
echocancel=no
language=fr-ca
context=ISDN-PROVIDER-10366283154eca8b328eef2-incoming
group=1
channel => 1-2


Thanks for help