Am 2013-11-22 um 18:30 schrieb Alexander Wagner <[email protected]>:
> If it's only about serving the bibliographic data I'd go for OAI-PMH as > it makes a smaller footprint and scales way better than dumping. Just > set your master to expose the collection as OAI-PMH and harvest it from > your slave periodically. Depending on the nature of your project you'd > most likely want to have the OAI-Server anyway for visibility reasons. Thank you for the idea, since I’m not familiar with OAI, I didn’t give consideration to that possibility. Visibility is not one of our goals... >> - after that, I change the server URL in all records (like in >> http://invenio-software.org/wiki/HowTo/HowToChangeSiteUrl) > > This, however, would not be necessary if you harvest. Additionally, > harvesting has the charming aspect that you need to transfer only the > diff and not the whole bunch. (IMHO dbdumps are possible but not a real > solution. It could be that I differ here a bit from CERN usecases ;) MySQL replication also just ‚diff’s, rsync the same. >> Doesn’t have anyone experiences with multi server setups to share? > > Our partners just harvest JuSER every night. Works perfectly especially > as you can harvest MarcXML right away so no transition formats necessary. Ok, sounds good. >> Isn’t it possible to save media files only with relative paths (since those >> are the same on all servers)? > > I don't get that. If they are all on the same network, why the fuss with > the laptop? However, you could just not include the full texts and let > the URI point to your master. (However, then I still don't get the laptop.) Ok, I must explain. I’m setting up Invenio for the eBilim project of University of Central Asia (see http://www.ucentralasia.org); while UCA will bring higher education to far off regions, eBilim is about general public education - since the breakdown of the Soviet empire, public education is in disorder. A bus with a digital library is regularly visiting villages in mountainous regions of Kyrgyzstan where there’s no internet access. This bus has a few workplaces for reading and watching videos. People can copy media to their own devices and teachers to their school’s computers, or print a few pages. (They have also physical media, esp. for kindergarden and grammar school ages, but not much. And there will also be events like newsreel cinema and workshops.) These workplaces as well as the server are laptops (they must cope with power breakdowns or run on generator power at all). The UCA staff in Bishkek (administration headquarter) fills our „public“ server (https://invenio.cerebrale.net); the media is partly Public Domain, but mostly made available by other projects, NGOs, newspapers etc. There’s not too much public educational media available in Russian, much less in Kyrgyz or the minority languages (e.g. Uzbek, Dungan, Uighur). They may occasionally hurt copyright, since it is not very respected in the region at all, or their license for media might be restricted to that one purpose, so we’re not so much interested in visibility. The „bus server“ is then sync’ed to the „web server“, as soon as the bus has proper internet access (normally probably only after each journey). Maybe Invenio is overkill for the main use, since we don’t need most(?) of the huge set of functions, but it seemed the most flexible and sustainable solution (and I know & like Python, but know & hate PHP and don’t know enough Java to adapt the other candidates). The administration didn’t decide yet about software systems for running the university (they plan to start in 2016), so our pilot project is also a test case for the main library database to be. Greetlings, Hraban --- http://www.fiee.net https://www.cacert.org (I'm an assurer)
