This is interesting. Thank you both. We're set on Docker for now, but these open up options if things don't go our way.
On Thu, Nov 29, 2018 at 4:02 PM Bram Vanroy Bram.Vanroy@ugent.be wrote:
Hi
Please don’t be afraid to use hundreds or even thousands of databases with BaseX. Colleagues of mine and I wanted to make querying huge (+500M tokens) treebanks (large, parsed corpora) faster. We preprocessed our data and ended up with millions of databases that were all served with a single BaseX instance. You can read about it here:
*Querying large treebanks : benchmarking GrETEL indexing*
Vanroy, Bram, Vandeghinste, Vincent and Augustinus, Liesbeth COMPUTATIONAL LINGUISTICS IN THE NETHERLANDS JOURNAL. 2017. 7 p.145-166
Bram
*From:* BaseX-Talk basex-talk-bounces@mailman.uni-konstanz.de *On Behalf Of *France Baril *Sent:* Friday, November 23, 2018 10:18 *To:* BaseX basex-talk@mailman.uni-konstanz.de *Subject:* [basex-talk] BaseX and docker
Hi,
For one of my projects, we want to set up BaseX as part of an online application. I was wondering if anyone has experience installing BaseX in a Docker setup and if so how things got split up.
In our case, each client will have many database, we're afraid that if we use a single baseX installation, it will have too many attached dbs.
So we were thinking that we could use 1 application but save1 data/repo per client in their own docker unit , then again, since .basex can only link to 1 repo/data, it doesn't seem possible. If you consider 1000 clients with an average of 30 (1 per language) database each, would you use a single installation and manage security at the db level with all dbs for all clients in the same setup or would you suggest 1 full instance (server
- set of db) per client?
Links to any case study would be useful too.
--
France Baril Architecte documentaire / Documentation architect france.baril@architextus.com