dying gasps & DLM

[sevenlayermuddle]

I decided to download a copy of G.992 for bedtime reading. One of the first things to catch my eye was the wonderfully named 'dying gasp' message.  This seems to be a mechanism wherby the router, in the event of power-loss, is supposed to retain (say, in a big beefy capacitor) just enough energy to communicate the fact to the DSLAM in a 'dying gasp' message.

It set me to wonder,  does/should DLM take notice of 'dying gasp' messages, and so tell the difference between frequent power-offs and other perceived instability?  Moreover, should it actually be safe to power on/off a router as often as you like as long as it's working properly (sends the dying gasp messages)?

From what I recall, the one time I got penalised by DLM I'd been pulling the plug on the modem lead rather than powering off the router. 

I'm afraid I don't feel like volunteering to try this out, so don't ask 

[roseway]

I'm sure that DLM does take notice of dying gasp messages, but I'm not sure that routers send those messages if you just switch them off. USB modems send dying gasp messages because they are shut down in an orderly manner when the PC is shut down. I rather doubt if regular domestic routers have big enough capacitors to keep them running long enough.

I'd be delighted to be proved wrong though.

[sevenlayermuddle]

I rather doubt if regular domestic routers have big enough capacitors to keep them running long enough.

I share your sceptisism but from g.992...

from g.992 9.2.5.4
The ATU-R shall have the ability to detect when the electrical power has been shut off...

.. and send dying gasps.   Note that's a 'shall', not a 'should' or a 'may'.   'ATU-R' refers, I believe, to the customer premises ADSL transceiver unit.

I suppose, ultimately, it only needs to store a few milliseconds of power.  But I'm still not volunteering to try and prove it.

[kitz]

From what I can recall..  the dying gasp is in the event of  power failure.
Pulling the modem cable out of the phone line does not send the dying gasp message.

>> Moreover, should it actually be safe to power on/off a router as often as you like as long as it's working properly (sends the dying gasp messages)?

lol I wouldnt like to try it..  from what has been seen many times over,  powering cycling the router certainly does appear to fool the DLM.

I'm assuming there must be something specific about say a power-cut that allows the router to be able to send the dying gasp message, that doesnt occur when the router is physcially powered down... and the "arrrrrrrrgh Im dying" message doesnt gets sent. (either that or BTs dslams choose to ignore the dying gasp!)
 
Finally... Pulling the phone line would obviously stop any dying gasp message getting through to the dslam.

[sevenlayermuddle]

lol I wouldnt like to try it..  

Damn, I though I'd found a volunteer   

Finally... Pulling the phone line would obviously stop any dying gasp message getting through to the dslam.

Absolutely, I only mentioned it in the context that my own raised target last year could probably be attributed to pulling the phone line, and so it doesn't prove (or disprove) that multiple power-cycles would have had the same result.

Seriously, I'm inclined from now on to make sure I power off the router before disconnecting the line.  It may not do any good, but I'll bet it does no harm?

[kitz]

aha - I only mentioned pulling the phone line as a warning to others that this is obviously the worst thing you can do.

Shutting down the router gracefully - like you say - theres no guarantees..  but it cant harm.