Answer the question
In order to leave comments, you need to log in
Sharding tokudb across openvz containers
Good afternoon. I would like to hear criticism of the idea.
A network application is being written, where database write operations will significantly dominate reading. There was a following thought. Let's take a few guys. We optimize one of them for recording, for example, take ssd. We raise openvz for everyone. We limit the size of the database to a size slightly smaller than allowed by the free license- 50 GB. As soon as the size has reached the limit, we inform clients (who also work in the server part of the application) that this container is no longer available for writing. After the end of the last recording, we transfer the container to the node for storage. We start a new container for recording and so on in a circle. As a result, we have a very fast recording and almost infinite storage for data. Clients working with the database are planned to be written in erlang. When a request for reading is received, it is simply sent in parallel to all storage containers, and, of course, to containers with working slaves for write nodes.
Answer the question
In order to leave comments, you need to log in
Looking at your description, I wonder why not buy? Why not use free mongu?
Didn't find what you were looking for?
Ask your questionAsk a Question
731 491 924 answers to any question