Finding some feedback on this and trying to move forward with solid design for my next project. Imagine 5,000,000 individual images with Lo-Ridge preview images; Each group of pictures belongs to the gallery, which belongs to a person, the number of people is X. A special version of Flickr, to say so.
To view these photos, they are placed on a remote host with the web interface. There is also a desktop application to go with, where you can upload images and enter gallery details on this remote host. The desktop app speaks about the preview image and information about each gallery and person, a desktop app that is synchronized with Flickr.
I have to decide on two things: Remote host for desktop app and backend storage of local storage is intended for a Windows environment, so I was thinking that SQL Server Express would be a good fit, But this project has grown quite a bit and it can only work for the desktop end.
Remote (web) server can be Windows or Linux, PHP or Net. Net - I do not care as much as the technology fits. The question is that the best way to store all the data on the web server is so easy that it can easily be indexed, accessed faster, and, most importantly, easily backed up and in the event of a disaster Could be restored. I'm not worried about server configurations or disk space at this time, as long as the database solution supports some of the cloud computing scenarios.
I am thinking that no-SQL backend is most understood, instead of storage images, galleries, and users' NO-SQL as a 'article' rather than 'rows' More likely to be on Flipside, Flickr has been advertised as using MySQL ...
Perhaps this is a more existential question, which is a real coding question, but I have a better group do not know ! After managing stock photo site with more than 5 million photos, I can say that MySQL is definitely a viable option. Backup is easy if you use replication, just stop the slave, copy it and then back it up.
MySQL full text search is not very good and may be slow, so you can check the solar-based Lusen engine. Elastic search is also a good option for scalability.
If you are unsure about your data structure, then there may be a good solution like Mongodibi, but there are limitations on the aggregation (~ 10,000 records) in Mongo, so keep this in mind. But this is one of the easiest setup.
Anything can work in the cloud, so I think the requirement is an important issue. You can set anything on EC2.
Comments
Post a Comment