Creating a web application (Django in my case, but I think the question is more common) that the cluster of workers Administering is doing queued jobs, there is a need to track progress of each work.
When I have used database updates (PostGreysQL) in this case, it seriously hits the database performance, because each update creates a new row in a table, and my case Removes obsolete rows only by removing DB in it every 10 days DB is required to run and report progress every 10 days (and this means that working with the system For all the employees staying away from the front, at a very slow pace)
Because progress information is not important, that is, it does not happen continuously, how will you update progress with jobs without using the overhead database? There are 30 employees in the server, each one or two jobs together, 1 front end server which provides web applications to users, and 1 database server.
is a package called a fast server for key-value recovery Sets. It is used by large clustered sites like Wikipedia.
This lets you share constantly-changed data around your cluster without the DB overhead.
Comments
Post a Comment