Specify Web Portal (Version 2.0)

The Specify Collections Consortium is funded by its member institutions. The Consortium web site is: http://wwww.specifysoftware.org

Specify Collections Consortium
Biodiversity Institute
University of Kansas
1345 Jayhawk Blvd.
Lawrence, KS 66045 USA

Developer Instructions

After completing these instructions you will be able to run Specify Web Portal 2.0.

These instructions are for deployment on a server running Ubuntu. An export file for a single collection is required for setting up the Specify Web Portal. This can be accomplished using the Schema Mappging tool tool within the Specify 6 application together with the stand alone Specify Data Export tool.

Install system dependencies

Python 2.7
Nginx webserver
JRE for running Solr
GNUMake
Unzip utility
cURL utility
python-lxml

Installation Instructions

These instructions illustrate the fewest steps needed to install the web portal. For updating existing installations, se the 'Data Only Updates' section below.

Clone the Web Portal 2.0 repository by clicking the green button (Clone or Download) at the top of the page and unpack this repository on your server.

This will install Solr on the server.
Use the Specify Data Export tool to create a Web Portal export zip file (see the Specify 6 Data Export documentation) for each collection to be hosted in the portal. If aggregated collections are desired, replace the single colleciton with the aggregated collections file after the initial Web Portal installation.
Create a specify_exports directory in the web portal directory and copy the zip files (step 2) into it. The copied files should be given names that are suitable for use in URLs; so no spaces, capital letters, slashes or other problematic characters. E.g. kufish.zip
Build the Solr app: make clean && make.
Copy the Solr core to the Solr installation:

[CORENAME] - the name of the exported archive without the file extension. (E.g. kufish)

[SOLRVERSION] - the version of Solr. (E.g. 7.5.0)

    mkdir solr-[SOLRVERSION]/server/solr/[CORENAME]
    cp -r build/cores/[CORENAME]/core/* solr-[SOLRVERSION]/server/solr/[CORENAME]
    cp build/cores/[CORENAME]/web.xml solr-[SOLRVERSION]/server/solr-webapp/webapp/WEB-INF/web.xml 
    # Only necessary for the first core.

Restrict access to the solr admin web page. This can be done in solr 7.5 by editing solr:

[SOLRVERSION]/server/etc/jetty-http.xml

In the ServerConnector section replace:

<Set name="host"><Property name="jetty.host" /></Set> with <Set name="host">127.0.0.1</Set>
Start solr solr-[SOLRVERSION]/bin/solr start

When completing this step the following warnings may be issued and can be safely ignored:

*** [WARN] *** Your open file limit is currently 1024.

It should be set to 65000 to avoid operational disruption. If you no longer wish to see this warning, set SOLR_ULIMIT_CHECKS to false in your profile or solr.in.sh *** [WARN] *** Your Max Processes Limit is currently 63590. It should be set to 65000 to avoid operational disruption. If you no longer wish to see this warning, set SOLR_ULIMIT_CHECKS to false in your profile or solr.in.sh Waiting up to 180 seconds to see Solr running on port 8983 [|] Started Solr server on port 8983 (pid=15422). Happy searching!
Import the csv data: curl 'http://localhost:8983/solr/[CORENAME]/update/csv?commit=true&encapsulator="&escape=\&header=true' --data-binary @build/cores/[CORENAME]/PortalFiles/PortalData.csv -H 'Content-type:application/csv'

When completing this step you may receive output similar to the following:

{

"responseHeader":{

"status":0,

"QTime":1153}}

Move the built webportal to convenient location:

[WPPATH] - the directory where the web portal fiels reside. (E.g. /home/specify/webportal)

mv build [WPPATH]
Create the Nginx configuration file to serve the portal:

sudo nano /etc/nginx/sites-available/webportal.conf

server {
    listen 80 default_server;

    rewrite ^/([^/]+)/select /solr/$1/select;

    location ~ ^/solr/([^/]+)/select {
      proxy_pass http://localhost:8983;
    }

    location / {
             root [WPPATH]/html;
     }
}

Remove the default Nginx site and enable the portal site:

sudo rm /etc/nginx/sites-enabled/default
sudo ln -L /etc/nginx/sites-available/webportal.conf /etc/nginx/sites-enabled/

Restart Nginx: sudo systemctl restart nginx

The Web Portal is now running and can be tested.

The Web Portal can be customized by editing the settings in the settings.json file found at here: [WPPATH]/html/corename/resources/config/settings.json

Data Only Updates

If the fields used in a portal are unchanged and only data is being updated, delete the current contents of the solr core with:

curl 'http://localhost:8983/solr/hollow/update?commit=true&stream.body=<delete><query>*%3A*</query></delete>' (You will probably need to add/edit a requestParsers block in the solr-[SOLRVERSION]/server/solr/[CORENAME]/conf/solrconfig.xml file for the core. Add it to the requestDispatcher block:

<requestDispatcher>
    <requestParsers enableRemoteStreaming="true"
                enableStreamBody="true"
                multipartUploadLimitInKB="2048"
                formdataUploadLimitInKB="2048"
                addHttpRequestToContext="false" />


    <httpCaching never304="true" />
</requestDispatcher>

(See http://lucene.apache.org/solr/guide/requestdispatcher-in-solrconfig.html)

Then use the curl CSV import command above to add the new data.

Schema Definition Updates

In this case, you will need to stop Solr (solr-[SOLRVERSION]/bin/solr stop), remove the cores to be updated from your Solr server directory, and follow all the installation steps besides the HTTP configuration.

Web Portal Application Updates

If a new version of the Web Portal is being installed it will be necessary to perform step 8 after building