Summary:	DAHLIN-00080: "Yellow" (RAI) alarm generation for E1 broken when in CRC4 mode
Reporter:	Guillaume Knispel (gknispel_proformatique)	Labels:
Date Opened:	2009-02-07 19:10:42.000-0600	Date Closed:	2009-06-26 15:01:15
Priority:	Minor	Regression?	No
Status:	Closed/Complete	Components:	wcte12xp
Versions:		Frequency of Occurrence
Related Issues:
Environment:		Attachments:
Description:	Under some conditions the RAI generation is broken when CRC4 is activated. When this happens the generated alarm indication is periodically pulsed high during something like few hundreds of ms with a period that I have not been able to measure precisely but which is around one or few seconds. In other cases the RAI is not generated at all while it should. What should happen is a continuous emission of RAI. (BTW, completely unrelated issue; the way Dahdi handles alarms has absolutely nothing to do with ITU-T or ETSI specifications, but I guess most people don't care, although this could probably be troublesome for homologation if still necessary in some areas) ---- Sufficient steps to trigger the problem, but I don't know whether they are necessary: 1. Cold boot a system containing a disconnected Wildcard TE122 configured like that: span=1,1,0,ccs,hdb3,crc4 bchan=1-15,17-31 dchan=16 (and load the module and start the span with dahdi_cfg) 2. Connect it to a started compliant E1 interface configured like that: span=1,0,0,ccs,hdb3 bchan=1-15,17-31 dchan=16 3. The TE122 in CRC4 mode switch between BLUE RECOVERY alarm and RECOVERY alone around twice per second, and correctly continuously maintain the RAI high. 4. The TE122 in CRC4 mode continues to behave correctly even when disconnecting and reconnecting the interface once or multiple times, regardless of how fast it is disconnected and reconnected. 5. dahdi_cfg multiple times (quickly) while observing the status of the received RAI on the other system, until it stops to be continuous. 6. The RAI should now be (incorrectly) pulsed (or maybe completely absent). 7. The TE122 in CRC4 mode will now continue to behave incorrectly when connected to a non CRC4 E1, either pulsing or not generating the RAI at all when it should be continuously generated. Disconnecting and reconnecting the cable does not help. 8. Connecting the TE122 in CRC4 mode to an E1 also in CRC4 mode make it works correctly again: once the alarms have disappeared, reconnecting to a non CRC4 line correctly result in continuous generation of RAI (we are now back in step 3...)
Comments:	By: Guillaume Knispel (gknispel_proformatique) 2009-02-14 19:10:44.000-0600 I've been able to do more checks about this issue and when pulsed, the RAI is precisely sent when the TE122 is in BLUE alarm. (previously I thought it was pulsed at a slower rate, but this was due to a not yet finished debouncing code I was writing -- this time i've been able to observe the raw RAI with no debouncing) By: Russell Bryant (russell) 2009-06-26 15:01:15 Please report this issue to Digium technical support. That way, it can be reported and prioritized internally and get the attention of the appropriate people. Thank you!