Eclipse Community Forums
Forum Search:

Search      Help    Register    Login    Home
Home » Archived » OHF » PDQ limit
PDQ limit [message #40460] Wed, 23 January 2008 19:41 Go to next message
Jesse Pangburn is currently offline Jesse Pangburn
Messages: 166
Registered: July 2009
Senior Member
Hi,
I'm using the bridge and testing PDQ against some PIX managers and have
run into a weird thing. A certain PIX manager returns something like 20
entries when you search for "chip moore" because they have the same
patient in every PIX source system's local assigning authority. While
watching the bridge process this request, I learned about "continuation"
(which for those as new to this as me, it bascially means you get part of
a result set and a pointer to the next part- you keep making requests
until you get all the data).

So I had a chance to play with the parameter to limit the size of the
result set and am thinking that someone is misunderstanding its use. When
I set the value to "5" and query against them, I see the bridge retrieving
the various patients in blocks of 5- yet it still retrieves all the
patients instead of stopping at the first 5 as I would have expected. So
I set the value to "3" to see if the block size of 5 was just coincidence,
it was not and I found that the requests returned blocks of 3 yet still
retrieved all the patients.

My conclusion is one of the following 3 cases must be true:
1. I'm misunderstanding this parameter :-) entirely possible
2. The bridge should take the first block of "3" or "5" in this case, not
do the continuation thing for the rest of the result set, and just return
the specified number of records.
3. The PIX manager should not be sending the continuation pointer (or
whatever the heck you call it) and encouraging the client to keep getting
more of the result set.

Anyone have an idea as to which case is true?

thanks,
Jesse
Re: PDQ limit [message #40648 is a reply to message #40460] Sat, 26 January 2008 11:33 Go to previous message
Sondra Renly is currently offline Sondra Renly
Messages: 39
Registered: July 2009
Member
Hi Jesse,

By using the bridge, there are certain short cuts that have been taken
to make adoption as quick and easy as possible. This is one of them. If
you were using the plugins directly, you have to request the next
continuation. The bridge (for now) does this automatically for you.

So for a bridge user, specifying a record limit is encouraged to ensure
the server can respond in a timely manner. In fact we added in the
plugins themselves a default of 10 records to avoid a mass return from
the server.

Please feel free to write a bugzilla enhancement request to extend the
bridge API in a future version.

Sondra


Jesse Pangburn wrote:
> Hi,
> I'm using the bridge and testing PDQ against some PIX managers and have
> run into a weird thing. A certain PIX manager returns something like 20
> entries when you search for "chip moore" because they have the same
> patient in every PIX source system's local assigning authority. While
> watching the bridge process this request, I learned about "continuation"
> (which for those as new to this as me, it bascially means you get part
> of a result set and a pointer to the next part- you keep making requests
> until you get all the data).
>
> So I had a chance to play with the parameter to limit the size of the
> result set and am thinking that someone is misunderstanding its use.
> When I set the value to "5" and query against them, I see the bridge
> retrieving the various patients in blocks of 5- yet it still retrieves
> all the patients instead of stopping at the first 5 as I would have
> expected. So I set the value to "3" to see if the block size of 5 was
> just coincidence, it was not and I found that the requests returned
> blocks of 3 yet still retrieved all the patients.
>
> My conclusion is one of the following 3 cases must be true:
> 1. I'm misunderstanding this parameter :-) entirely possible
> 2. The bridge should take the first block of "3" or "5" in this case,
> not do the continuation thing for the rest of the result set, and just
> return the specified number of records.
> 3. The PIX manager should not be sending the continuation pointer (or
> whatever the heck you call it) and encouraging the client to keep
> getting more of the result set.
>
> Anyone have an idea as to which case is true?
>
> thanks,
> Jesse
>
Re: PDQ limit [message #584559 is a reply to message #40460] Sat, 26 January 2008 11:33 Go to previous message
Sondra Renly is currently offline Sondra Renly
Messages: 39
Registered: July 2009
Member
Hi Jesse,

By using the bridge, there are certain short cuts that have been taken
to make adoption as quick and easy as possible. This is one of them. If
you were using the plugins directly, you have to request the next
continuation. The bridge (for now) does this automatically for you.

So for a bridge user, specifying a record limit is encouraged to ensure
the server can respond in a timely manner. In fact we added in the
plugins themselves a default of 10 records to avoid a mass return from
the server.

Please feel free to write a bugzilla enhancement request to extend the
bridge API in a future version.

Sondra


Jesse Pangburn wrote:
> Hi,
> I'm using the bridge and testing PDQ against some PIX managers and have
> run into a weird thing. A certain PIX manager returns something like 20
> entries when you search for "chip moore" because they have the same
> patient in every PIX source system's local assigning authority. While
> watching the bridge process this request, I learned about "continuation"
> (which for those as new to this as me, it bascially means you get part
> of a result set and a pointer to the next part- you keep making requests
> until you get all the data).
>
> So I had a chance to play with the parameter to limit the size of the
> result set and am thinking that someone is misunderstanding its use.
> When I set the value to "5" and query against them, I see the bridge
> retrieving the various patients in blocks of 5- yet it still retrieves
> all the patients instead of stopping at the first 5 as I would have
> expected. So I set the value to "3" to see if the block size of 5 was
> just coincidence, it was not and I found that the requests returned
> blocks of 3 yet still retrieved all the patients.
>
> My conclusion is one of the following 3 cases must be true:
> 1. I'm misunderstanding this parameter :-) entirely possible
> 2. The bridge should take the first block of "3" or "5" in this case,
> not do the continuation thing for the rest of the result set, and just
> return the specified number of records.
> 3. The PIX manager should not be sending the continuation pointer (or
> whatever the heck you call it) and encouraging the client to keep
> getting more of the result set.
>
> Anyone have an idea as to which case is true?
>
> thanks,
> Jesse
>
Previous Topic:Human Requestor not included in XDS.a Doc Consumer Audit Record
Next Topic:NA Connectathon updates to OHF / HIMSS Prep / EU Connectathon Prep
Goto Forum:
  


Current Time: Mon Jul 28 20:37:56 EDT 2014

Powered by FUDForum. Page generated in 0.02223 seconds