SSDS

  • Comments posted to this topic are about the item SSDS

  • That's an interesting idea - replicating out all your lookup data to reduce the load on the "authoritative" server - basically this would need the same approach as implemented in the Domain Name System (Time-To-Live attributes etc.)

    However, you would need some additional controls to either ensure integrity between the lookup data and the transactional data if you wanted to perform any modifications. Maybe you could delegate the responsibility by giving users some way of making an informed choice between speed and accuracy of the data?

    No, on second thoughts, I think that this approach would have to be for queries only...and I'd probably stick with one of the more robust replication models if I needed to scale out database load.

    Malcolm
    DB Ghost - Build, compare and synchronize from source control = Database Change Management for SQL Server
    www.dbghost.com

  • I dont think relying on the cloud is much worse than relying on a server. If you're counting on that solution to work and it fails, life is hard for a variable amount of time depending on how much thought you've put into it. Pay as you go is an interesting model, but depending on the formula it could be challenging to humorous; charge for space, suddenly developers are interested in using the right data type and normalizing, charge per query and we'll see a lot of client side caching, charge per read and we'll see a sudden interest in how to index properly!

  • None of the links in the editorial point directly to SSDS, but this one does:

    http://www.microsoft.com/sql/dataservices/default.mspx

  • Whoops and thanks for the link, William.

    I think that this can be a good solution for smaller companies. If you can't afford high bandwidth or distributed servers, this is one good way to perhaps improve performance. As I mentioned, I wouldn't use it transactionally though, maybe just lookup or reference data I have elsewhere.

    Or even perhaps to duplicate some other data that a mobile device might need to pull down and have trouble getting from the office. you could easily (I assume) replicate data from your main system using the API to the cloud.

    The charge by use is interesting. It would encourage more efficient designs!

  • this will might be a good way as a online archive solution

Viewing 6 posts - 1 through 5 (of 5 total)

You must be logged in to reply to this topic. Login to reply