TLDR: What are the embedded db systems that support replication with no server? Ideally functional with .Internet

Background: Previously I had been requested to produce several MS Access databases that are increasingly being used over the network and gratifaction is quite slow. I've now been requested to update these databases to include newer and more effective features and improve performance. Ideally I must possess a server based after sales like MS SQL Server, regrettably it does not seem like which will happen. I'm searching for some type of database I'm able to use where I've got a master copy on the network share and client copies can replicate with this particular master with no master running on the server. Much like MS Access replication. I'm preferred dealing with .Internet languages like C#. If everything else fails I'll most likely need to try MS Access replication, however i was wishing for something a bit more robust.

Added Information: I'm not within the IT department inside my company and for that reason do not need a server system. We are attempting to obtain a single database that people might have admin privileges over with an existing SQL Server, but it's a constant fight. I'm searching for options that offer peer to see replication to beat the performance issues I'm going through with MS Access. I had been also searching for something which might have more support insInternet since Access does not use Linq or EF.

You could utilize SQL Server Compact. Earlier versions support synchronization from the central store however in 4. you could utilize something similar to this Snapshot Synchronization to synchronize the information.

The majority of solid replication architecture derive from Document-oriented databases.

Within the .Internet world, you should check NoSql databases : peformance &lifier scalability. Check Masterless Replication in NoSQL solutions.

Also, maybe you will not choose this architecture this year but Lotus Notes/Domino provides strong replication accross servers and clients (by having an offline mode).

However, the databases aren't SQL-based even when you are able to in some way "synchronise" all of them with a SQL server.

Without having a distinctive reason for admission to your computer data (just like a single database server) you instantly have to handle lots of replication hassle : concurrency, simply to title one.

It may sound much like your issue is a performance one, and also you tought about replication being an avenue, shall we be held right? If you're able to acquire a good distributed performance without replication, will it fit?

You will find No-SQL tools (like JoeBilly stated) like Memcached or even the Sterling framework. The concept behind is you handle in-memory teams of objects, without persisting these to a normal SQL database. But you may also persist them anyway diversely (the frameworks handle some situations for your).

If you're to revert to MS-Access, then the free SQL Server Express could be better for me. But this still takes a server (based on volume, a normal desktop could do). Another database engines handle replication (I believe PostgreSQL (http://www.postgresql.org/) is a - and it is free).

You may could elaborate in your scenario and constraints?