Coming Out of the Cloud

  • Comments posted to this topic are about the item Coming Out of the Cloud

  • I find that a major challenge is getting fast running local tests in place.  In AWS there is the moto library that mocks may of the objects covered by the boto library.  Some AWS services are covered very well, such as S3, whereas others implement only a handful of metadata calls.

    By the nature of the beast it may not be feasible to have a local mock of something that is a distributed framework.

    I found that when using AWS RDS for PostGres the main differences were in the permissions that were granted and a PostGres extension.  An alternative that vendors could provide is a simple script to provide cloud level privileges on a chosen DB.  That is, the DB is constrained as it is in the cloud.

    When SQL Server Developer Edition was 1st released it was available at a price.  Later on it became a free product.  This meant that I could have a throttled back Enterprise Edition on which to test and learn.  Being able to have local components reduces learning friction and inertia.

    For Azure you have Azurite  to provide a local equivalent of Azure Blob Storage.

     

     

     

  • Good questions. We're in the process of migrating several databases to the cloud. Dev and test environments, first. But just like you said, Steve, it is mandated by management and it's all lift-and-shift. I don't know why, as I'm not involved in those discussions. My guess is there are two reasons for doing it this way. The DBAs and system people are just more comfortable with an architecture they're familiar with (VMs, etc.), so IaaS gives them what they want. But also, there's an extraordinarily powerful desire not to modernize any apps. Those two things put together means we're just lifting and shifting database environments, then simply updating connection strings. I'm concerned that at some point by next year or sooner, the cost of leaving all software based upon decades old architecture, plus having databases in the cloud, will only result in a mad rush to get everything back on-prem, without really having tried to modernize our apps and processes.

    Kindest Regards, Rod Connect with me on LinkedIn.

  • David.Poole wrote:

    When SQL Server Developer Edition was 1st released it was available at a price.  Later on it became a free product.  This meant that I could have a throttled back Enterprise Edition on which to test and learn.  Being able to have local components reduces learning friction and inertia.

    Except Dev Edition = Standard, which causes issues, and != Azure SQL Db, which is a bigger issue. I want a trace flag to make it behave like those.

  • "..I'm especially surprised how many have performed lift-and-shift migrations to IaaS services after a mandate by management. I'm also not surprised that many customers find they're spending too much, and that both Azure and AWS realize they need to help customers spend less before they lose them.."

    Perhaps the main reason why your cloud hosted services would exceed your original expectations is because your footprint of applications and volume of data has expanded by an order of magnitude over the past few years - what with scale out replicas and Log Analytics and Blob storage and TB sized OneDrive accounts for every employee. But once you settle into it - can all that infrastructure be shifted back to an on-prem data center for less cost?

    "Do not seek to follow in the footsteps of the wise. Instead, seek what they sought." - Matsuo Basho

  • Very good point, Eric. Certainly, one I've not thought of before. I've not been here a decade, but I'm sure you're correct in that our application footprint has grown by at least one order of magnitude. Perhaps even more than that.

    Kindest Regards, Rod Connect with me on LinkedIn.

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

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