Transcript Higgs 02 03 HermesLite

GAIA RESOURCES
Experiences in mobilizing biodiversity
data – the story of HermesLite
Tim Carpenter and Piers Higgs
P088; Presented in Canberra, 27th March, 2008
GR000: Presented in Fremantle on 20th October, 2008
INTRODUCTION
• Motivation
• HermesLite
• Overview
• Architecture
• Implementation
• Implementation at the WA Museum
• Issues
• Directions
MOTIVATION
• Based on our experiences with Museums:
• Cost had to be minimal
• No capacity to serve data
• Need minimal infrastructural change
• Remote management necessary
• Funding for WA Museum (WAM) was made
available through the Australian Biological
Resources Study (ABRS)
• Under existing support arrangements to the WAM,
Gaia Resources undertook the actual work
HERMESLITE
Overview
• HermesLite is an application designed to help
institutions serve data - in accepted formats and
standards - where they don’t have the capacity
or resources to do this themselves.
•
•
•
•
Open Source
Written in Java
Easy to manage within an institution
Can be remotely managed
HERMESLITE
Overview
• Deployable in two configurations:
• Data Retrieval mode: A liteweight, extensible,
configurable Java application to extract and
send data to…
• Data Receiver and Storage mode: Extensible
Java servlet application to receive, translate
and store the data.
• The Data Receiver and Storage mode can also
act as the Retrieval mode if databases are local.
HERMESLITE
Architecture
External Server
Internal Server
Internal
Collection
Database
HermesLite
Retrieval
mode
MapServer
HTTPS
XML
HermesLite
Receive &
Store mode
External
Database
TAPIRLink
HERMESLITE
Implementation
• Data Retrieval mode:
• Unpack distribution files onto server
• Create a configuration file for each database
• Set up Java Wrapper Service to run
appropriately
• Data Receiver and Storage mode:
• Install HermesLite web application
• Extend HermesLite classes
• Drop in handler JAR to process data
• Install other software, (e.g. TapirLink)
HERMESLITE
Implementation - Data Retrieval
• Configuration
via a simple
properties file
with
variables
HERMESLITE
Implementation - Data Storage
• Extend
HermesLite
classes to
process
received
rows against
chosen data
format or
standard
HERMESLITE
Implementation
• This can represent a considerable organisational
change in the way data is delivered
• Be prepared to guide your organisation through
this process
EXAMPLE
Western Australian Museum
• Considerable discussion, demonstration, etc
(“social engineering”) – still ongoing after four
months
• Selection statements for each database
(restricting “sensitive” data) – being tuned all the
time, curators define these
• Darwin Core format
• Restricted the fields being delivered
EXAMPLE
Western Australian Museum
Gaia Resource Bureau Server
WAM Server
WAM
Collection
Database
HermesLite
Retrieval
mode
MapServer
HTTPS
XML
HermesLite
Receive &
Store mode
Bureau
Database
TAPIRLink
• Production went live on 11th October, 2008
• Data services are IP restricted to OZCAM only
ISSUES
Two main issues struck during implementation:
• “Social Engineering”
• Life Sciences IDentifier (LSID) resolution
ISSUES
Social Engineering
• Are all parts of the organisation ready to publish
their data? Consider:
• Management?
• Data custodians (curators)?
• Information Technology staff/providers?
• Customers (both “big” and “little”)?
• Requires considerable cultural changes to
organisations, especially when revenue from data
provision
ISSUES
LSIDs
• A bureau service has issues with LSID resolution:
• Originating institution WAM
(www.museum.wa.gov.au)
• Bureau service – hosting and resolving – at
Gaia Resources (www.gaiaresources.com.au)
• Difficulties in getting access to DNS records
• Currently investigating opportunities to get around
this by using the OZCAM cache as the LSID
resolver
DIRECTIONS
Future Directions
• Discussions with GBIF and the ALA about
integrating with other provider toolkits
• Potential use within the Museum community in
Australia under the Bureau service model
DIRECTIONS
TAPIRLink
Collection
Database
Bureau Server
CSV
Collection
Database
HermesLite
Retrieval
mode
Collection
Database
HermesLite
Retrieval
mode
MapServer
HermesLite
Receive &
Store
mode
Bureau
Database
TAPIRLink
ACKNOWLEDGEMENTS
Would not be possible without:
• Funding from the ABRS
• Support from the WAM, both from management
and the curators themselves
• Assistance and advice from the Faunal
Collections Informatics Group members
MORE INFORMATION
For more information…
• Links and information available from the Gaia
Resources web site, including links to the
SourceForge project.
www.gaiaresources.com.au