[emstar-design] more questions on motenic...

[Karen Weeks]

I am just wondering why you say it doesn't get sent.  I see these packets 
going from the host to the mote without any ill-effects.  I will try 
testing next with ping though, possibly sending packets more quickly will 
cause the problem to manifest itself more quickly / reliably.

Oh, and in response to your previous email, I set the ACK timeout to a 
very conservative value.  With the ACK timeout set to the deafult value, I 
would sometimes (not always) get ACK timeouts.  I increased the value a 
lot to ensure I 
never got ACK timeouts.  After all, when I am running over a socket, no 
packets can get lost, so I didn't think this was a big deal.

So anyway, Avrora does seem to run (in general) a bit slower than a 
'real' mote, but I don't want people worrying that it is running a lot 
slower.

Karen

On Mon, 23 May 2005, Athanasios Stathopoulos wrote:

> On Mon, May 23, 2005 at 01:44:35PM -0700, Karen Weeks wrote:
>>
>> OK, that was what I was missing.  Is this an example of a data packet?
>>
>> mDtn: [usr3]TimeSynchM$sendDataToTree: SINK: Sending time packet with time
>> 19706
>> Mon May 23 13:38:00.248 2005: mDtn: [usr3]ESS_mDTNM$sendHandler:
>> DTN::SENDING PKT FOR 4
>>
>> If so, these packets do seem to be sent and ACK'ed successfully at least
>> some of the time (in other words, they do not consistently cause the
>> system to lock itself up).
>>
>
> yes, thats a packet sent from mdt that gets put in the motenic and yea that
> doesn't get sent. That's why I said the data protocol isnt working.
>
> Again, stop using mdtn! it just clutters the logs and gets confusing. Plus,
> if you're testing low level functionality (motenic) you don't want a complex
> high-level protocol (mdtn, multihop, all that stuff) to run on top of it...
>
> Use emstar ping
>
>
> -thanos
>