Node.js图像库的数据库解决方案

Question

I am developing a personal web-based aircraft photo gallery using Node.js, to show off my photos to friends/colleagues/etc. The server would host some hundreds or 1-2 thousands of these photos (as static files) at maximum, while their accompanying data would be stored in a database.

The website should not generate too much traffic (1-2 accesses only at a time). Its main feature would be a complex search "engine", capable of filtering and ordering photos by aircraft manufacturer, aircraft type, aircraft registration, airline, airport where the photo was taken, date of the photo, and numerous other aspects.

I have developed a relational model to serve these goals, to be implemented in a MySQL database:

                               +-------+
                               | Photo |
                               +-------+
                              N |     | N
               +----------------+     +--------------+
             1 |                                     | 1
  +------------------------+            +--------------------------+
  |Aircraft | e.g. "G-CIVL"|            | Airport | e.g. "Heathrow"|
  +------------------------+            +--------------------------+
             N |                                     | N
             1 |                                     | 1
   +-----------------------+              +-----------------------+
   | Model | e.g. "747-400"|              |  City | e.g. "London" |
   +-----------------------+              +-----------------------+
             N |                                     | N
             1 |                                     | 1
 +-------------------------+         +--------------------------------+
 |Model_family | e.g. "747"|         | Country | e.g. "United Kingdom"|
 +-------------------------+         +--------------------------------+
             N |
             1 |
+----------------------------+
|Manufacturer | e.g. "Boeing"|
+----------------------------+

Every photo entry would contain information about the location it was taken at (Airport - city - country "branch") and the aircraft shown in the photo (Aircraft - model - model_family - manufacturer "branch").

As I've mentioned before, the visitors should be able to filter photos by any columns of these tables, requiring different joins in the database query for different filtering options. Examples:

1. The user is looking for photos taken in a certain country. The SELECT query returning the corresponding image data should join the Photo, Airport, City and Country tables.
2. The user is filtering photos by a certain aircraft model. The SELECT query would have to join Photo and Model tables only.

The question is even more complex if the user is filtering photos by multiple (or even more) aspects at the same time.

For this reason, a complex application-level SQL query building algorithm would be needed, which would take the aforementioned table structure into account.

My questions are:

• Is the mentioned method (complex SQL string building algorithm) a good approach to the problem?

• Or should I just join all the tables together regardless of the user's filtering options? (seems to be a serious performance tradeoff for better code maintainability)

• Or would a NoSQL solution like MongoDB with a flatter (and denormalized) model more adequate for this task? Not having to set up different joins for different filtering options seems to be a huge plus. It is very unlikely for the stored data to be updated or deleted, therefore denormalization might not be a serious problem.

Thank you for your help in advance.

Answer 1

I haven't used MYSQL that much but in NoSQL ( Mongodb )

you can get data structure something like this and you can search based on every parameter

[
  {
    "_id": 0,
    "Airport": "Heathrow",
    "city": "London",
    "country": "United Kingdom",
    "Aircraft": "G-CIVL",
    "model": "747-400",
    "model_family": "747",
    "manufacturer": "Boeing",
    "photo_url": "http://www.dummy-url/dakshdajk8sdjkhs8ds8dh8sd8sdj8s.png"
  },
  {
    "_id": 1,
    "Airport": "Heathrow",
    "city": "London",
    "country": "United Kingdom",
    "Aircraft": "G-CIVL",
    "model": "747-400",
    "model_family": "747",
    "manufacturer": "Boeing",
    "photo_url": "http://www.dummy-url/dakshdajk8sdjkhs8ds8dh8sd8sdj8s.png"
  }
]