[OAI-implementers] resumptionToken Implementation

[Chris Wilper]

In Fedora, we use the following approach: The query is run once against the
database.  After iterating through the first chunk of results, the ResultSet
object is just put into Map, keyed by a unique string (which is also the
resumptionToken).  If no one comes back to continue getting results within
some period of time, the ResultSet is removed from the Map and close()d.  If
they do come back, we can immediately start doing next()s.

The advantage is that there's very little latency on subsequent requests in a
series of resumptions.... the drawback being that DB connections from our
connectionpool are kept open until someone comes back, or our resumption
token timeout kicks in.

- Chris

-----Original Message-----
From: oai-implementers-bounces at openarchives.org
[mailto:oai-implementers-bounces at openarchives.org] On Behalf Of Jeff Pearson
Sent: Tuesday, September 28, 2004 12:32 PM
To: oai-implementers at openarchives.org
Subject: Re: [OAI-implementers] resumptionToken Implementation

I guess I misstated my query last time. I understand the implementations as
defined in the spec; either create a data result cache and hit that or
regenerate the query each time. What I was wondering was which people were
choosing to implement and why.


_______________________________________________
OAI-implementers mailing list
List information, archives, preferences and to unsubscribe:
http://openarchives.org/mailman/listinfo/oai-implementers