parsing files (such as log)
Processing of the logs and uploading them to a database can be relegated to any language that has access to the system API and will be able to use file mapping in conjunction with monitoring changes in the file for quick reading new real-time data, then parsing with regexps and insert in database.
I would choose nodejs, but through the use of file mapping highly doubt it...
saving to the database(some database need?)
Logs are very well suited to NoSQL graph database
Neo4jor MongoDB, Cassandra, elasticsearch and others. You do not need to monitor the presence of columns in the table and update all record, I doubt if you're writing super complicated queries to the database, such as Neo4j solves this very quickly:
To select users who liked the actors who starred in the movie, which sounded saundtreki that were written by musicians that I like
Something like this:
MATCH (me:User {userId:123})-[:Like]->(musicants:User)-[:Author]->(s:Soundtrack)-[:Used]->(f:Film)<-[:Starred]-(actor: User)<-[:Like]-(u:User) RETURN u
(c)copiedcreating windowed applications, a simple conclusion from DB
Whatever you want, though the same node.js on the basis of nw.js or electron.
I would use node.js possible, things can be even through a socket.no io requests to the database to get new records (inserted in dB) from the client.