NoSQL DBMS for web server on VDS

V

Vladimir Chernyshev2011-02-12 15:42:34

NoSQL

Vladimir Chernyshev, 2011-02-12 15:42:34

Inspired by the answers to this question . It turns out that it is highly desirable for MongoDB to allocate a separate server, since it does not want to give up memory, which can be fraught with other applications.

In this regard, the question is which DBMS is better to install so that its appetites for memory can be limited. It is desirable to be as close as possible to Mongo, that is, a free scheme of objects / documents, but with their division into collections.

Thank you.

Upd.: OS - Debian 6.0, nginx+php-fpm+passenger+mysql

Reply

Answer the question

In order to leave comments, you need to log in

5 answer(s)

R

RuJet, 2011-02-12
@RuJet

CouchDB

T

TimTowdy, 2011-02-12
@TimTowdy

In addition to the aforementioned rubber, it is hardly possible to recommend something else that is ready for production and works on VDS.
In general, the benefits of NoSQL on VDS are highly questionable. NoSQL is in demand where unlimited horizontal scalability is needed. If there is no such need, MySQL / Postgres are perfect for almost any task. You should not chase fashionable buzzwords, each task has its own tool.

A

andoriyu, 2011-02-13
@andoriyu

CouchDB and so far only CouchDB. It is surprising that only Mongo's memory consumption surprised you, but the fact that there is no concept of node integrity and without replication there is no point in going further than development.

P

Puma Thailand, 2011-02-12
@opium

All the same, for the most part, nosql solutions are oriented to an insane load and, as it were, the issue of moving to a separate server or little memory is not worth it.
Can you try to rent some kind of server?
ntel® Core™ i7-920 Quad-Core 4 physical cores, 4 virtual (hyper-threading) 2 x 1500 GB HDD SATA2 (software RAID1) 24 GB DDR3
Installation 150 euros, rent 89 per month
can be cut into kwm, openvz, xenon virtualok and screw everything up.

F

funca, 2011-02-13
@funca

I read that PostgreSQL has advanced features for working with XML - this format is suitable for storing documents with an arbitrary schema. You can make selections by xpath, build indexes, and even rob ... separate fragments of documents. Maybe it is worth looking for a solution in this direction?