Hi all,
I have been toying with the process of implementing common LIS
applications (e.g. Vufind, Dspace, Blacklight. . .) on PaaS providers
like Heroku and Amazon Elastic Beanstalk. I have just tried out of
the box distributions so far and have not made much progress but was
wondering if
Dear Erik,
Bram Wiercx and myself have given a talk on how to put together a package to
install CollectiveAccess on Red Hat's OpenShift:
http://www.dish2011.nl/sessions/open-source-software-platform-collectiveacces-as-a-service-solution.
My students are currently happily playing around with
Erik,
I haven't tried it (recently) on PaaS providers, but I have on IaaS. The
AMIs I've created in association with start up scripts (if you're
interested in seeing those let me know, I'd have to look for them somewhere
or other) mean that the application automagically starts up on its own, all
Hi,
I've deployed Blacklight on both Heroku and Elastic BeanStalk.
Heroku is still a much better choice. The only issue I had was I
needed to make sure the sass-rails gem in installed in the :production
gem group and not just development.
I still have an issue of getting heroku to compile all
If you already have everything indexed in Solr elsewhere, a way to cut down
the BL slug size is to remove/ignore the SolrMarc.jar. It's pretty sizable.
-Sean
On 3/29/12 12:16 PM, Chris Fitzpatrick chrisfitz...@gmail.com wrote:
Hi,
I've deployed Blacklight on both Heroku and Elastic
Hey Sean,
Jah, I did that...my .slugignore is:
tmp/*
log/*
coverage/*
spec/*
koha/*
jetty/*
That dropped it down to 30 from ~50mb, so that's good .
(koha has some scripts wrote to pull from our ILS).
I think the slug size is a really minor issue. Heroku says under 25mb
is good, but over 50mb is
Hi Erik,
While my FRBR-Redis datastore is not yet a common LIS application, I develop
and run it on a Rackspace Linux VM with Django and Redis. I am in the process
of migrating this stack over to Heroku for a Library App development workshop
I'm leading in May.
Jeremy Nelson
Metadata and
Chris - where did you deploy your SOLR instance and did that create
any issues for deployment (other than ignoring files)?
Erik
On Thu, Mar 29, 2012 at 12:37 PM, Chris Fitzpatrick
chrisfitz...@gmail.com wrote:
Hey Sean,
Jah, I did that...my .slugignore is:
tmp/*
log/*
coverage/*
spec/*
Dear All
At Ghent Universty Library in Belgium we are re-evaluating our web strategy in
several internal workshops with faculty librarians. The theme of our next
meeting will be about discovery, and one-search-box integrated search
solutions. We see these days a lot of vendors and cloud
Like Chris, I've deployed Blacklight on Heroku, and this thread
(particularly Rosalyn's message) has gotten me to write up a quick
HOWTO on the Blacklight wiki [0].
For Solr hosting I've used both a VM that I run (on Slicehost) and EC2.
Mark
[0]
Neat! Thanks Mark,
Erik
On Thu, Mar 29, 2012 at 2:19 PM, Mark A. Matienzo m...@matienzo.org wrote:
Like Chris, I've deployed Blacklight on Heroku, and this thread
(particularly Rosalyn's message) has gotten me to write up a quick
HOWTO on the Blacklight wiki [0].
For Solr hosting I've used
Older 3.x versions of Blacklight may have put a solrmarc.jar inside your
app's ./config/SolrMarc. That may not be caught by your slug ignore.
This was an error, it was never meant to do that. If you have one in a
BL 3.x you should be safe to remove it.
Other than that, I'm curious what's
Hey Erik,
I used this AMI for solr =
http://www.lucidimagination.com/blog/2010/02/01/solr-shines-through-the-cloud-lucidworks-solr-on-ec2/
Note : You will have to change the schema and solrconfig files on the image...
b,chris.
On Thu, Mar 29, 2012 at 9:44 PM, Erik Mitchell mitch...@gmail.com
Oh, yeah...Mark's wiki page also points out that you'll need to change
to postgres for production. Very important.
From what I understand, the compiled slug is the app + all it's
dependancies. So, yeah, I guess since the gem does have the
SolrMarc.jar in there, it would have the jar in the slug
On 3/29/2012 5:05 PM, Chris Fitzpatrick wrote:
locally and push them rather than rely on Heroku to precompile them
(currently when I push, Heroku's precompile fails, so it reverts to
compile at runtime mode) if anyone has insight into this, please
lemme know...I believe having them compile
Hi Graham,
What we do in our federated search system, and have been doing for some few
years, is basically give the designer a choice of what options the user gets
for de-duped records.
Firstly de-duping can be of a number of levels of sophistication, and a many of
them lead to the situation
16 matches
Mail list logo
