[news.eclipse.technology.ohf] Re: Missing object ID in ATNA logs from br

[Sarah Knoop <seknoop@xxxxxxxxxx>]

Hi Jesse - this is a bug with NIST (at least it was ... unless something 
has changed in the IHE specifications - will investigate more at the EU 
connectathon this next week to be sure)

see -
http://ihe.univ-rennes1.fr/flyspray/?do=details&id=183

So this at least explains your ITI-16 problems ... possibly the ITI-21 
(sondra check?). For retrieve document set ... We caused some animated 
discussion on the ititech mailing list this fall regarding this. This 
transaction did not get updated by the audit CP and when you retrieve a 
document set, there was no specification on how many audit messages to 
even generate when you actually get a set of documents back from 
multiple destinations. So we did the best we could. We create one audit 
message per destination retrieved from. Problem is if we get more than 
one document back fro this destination - which documentUniqueID do we 
audit in this missing (singularly specified in the CP) audit parameter? 
So I think we give a list of these id's (should double check). The NIST 
audit record parser must not like this. So this is still an open issue 
in IHE, that we probably should write a CP for.

hope this clarifies, despite my awkward wording,
- Sarah

Jesse Pangburn wrote:
> I see in my ATNA logs via the NIST browser that some transaction types 
> are flagged and the problem is highlighted as an empty "Object ID" field 
> in the "Participant ObjectId 1" section.  I know we ran into something 
> like this in Connectathon and I know there was at least a partial fix 
> because I remember seeing the object id get populated after applying 
> some fix.
>
> At this point, with the March 28 build, I see the warning sign next to 
> the following transactions:
> ITI-43 Retrieve Document Set
> ITI-21 PD query
> ITI-16 Registry SQL Query
>
> The following are fine:
> ITI-18 Registry Stored Query
> ITI-17 Retrieve Document
> ITI-15 Provide and Register Document Set
> ITI-9 PIX query
>
> This is not an exhaustive list as it's just the ones I was playing with 
> in the last couple days.  Any idea what's going on?  Are those 
> transactions just not supposed to have an object id?  That seems 
> unlikely as 16 and 18 are very similar and one has it and the other 
> doesn't.
>
> thanks,
> Jesse