dmj

100% on defining infrastructure as code and not using the UI. I know I keep saying it, and others here are opposed to doing so, but it DOES make a big difference in terms of accountability, reproduciblity, testing, DR and just general reliability.

Functionality aside, Arbor seem to have a 'better' tech stack in terms of their ability to scale and be a modern web platform.

How would you know if the software that monitors the software is working or returning false positives? You could go on forever. Ultimately you have to either trust what you bought, or get something that works (ideally with the capability to monitor itself). It's one of the reasons I prefer network level monitoring rather than device level. As a first step, at least check the logs for when something does go wrong.

It probably depends on the team, their skillset and the school philosophy. There's not a right or wrong answer: it will work for some but not others.

This is also how Opensource projects get started, and companies form around them. Sounds like the mistakes your school made was not to share.

@miscbrah Looks like you found yourself a product manager for this project. Add these to the roadmap

Replace linux with windows in 2024. Radical.

That's literally what I would do to, perhaps native KVM.

Yeah there's a thread about (one of the Bromcom threads IIRC) how SIMS changed their contracts to prevent third parties accessing the database to drive out competition. Apparently the CMA say that's fine, but the DfE say hosted is the way forward. Both of those pieces of information should be enough to convince LA's and schools that SIMS is a dead end until they actually release their vaporware platform.

IT tech time in education is always cheaper than paying developers wages

is this actually in the spec? Could someone point me to it as I'd love to see the exam board justification. Just seems an odd choice when Computer science teachers are in a shortage they would choose a language with half a percent market share. Perhaps easier for teachers with NO programming experience to pick up - and fewer examples to 'copy' from the internet* * a major part of a professional developers time is spent on doing exactly this.

IME you've got to have a very good reason to introduce a breaking change, and I'm not sure you've even convinced yourself of the benefits.

Yeah it's mostly the same as it's always been in 'enterprise' Just be sure to use uptodate encryption. AD should distribute everything you need.

Sounds like it will break things for very little benefit.

I'm not keen on either of these methods TBH, this should be a straight API call. You shouldn't need to have a database to replicate another datasource, it just adds unnecessary duplication. I'd also advise against Mysql, if your php framework supports postgres it is a much better choice. Should probably send a link to the URL with the indicent, rather than emailing confidential data. Thinking data retention, GDPR etc. There are enough libraries in php which will give you SAML/OIDC for you Azure/Google/Directory that will handle users.

Purely out of interest, why do you keep referring to 802.1x as 'Modern' authentication? It's been around for 20yrs, I'm sure Windows XP allowed this and we certainly had it setup on Win7.

I was working at a Chromebook school where all the students had Chromebooks, but the teachers had to have windows just for SIMS. Ditching SIMS would have allowed us to ditch all of our MS licensing and allowed us to use lower cost computers. We could have cut about 80% of our entire IT budget (or spent it on educational resources). Unfortunately some of the SLT liked SIMS so we had to keep it.

Sounds like a caching issue than DB to be honest. The amount of places things are cached these days is crazy.

You can factor in the increased management costs of windows, which if done properly would likely preclude them in favour of simplicity anyway

Yeah, did spot that. I've seen some ITIL processes direct to CAB for standard changes if they have budgetary impact- so figured it was one of those.

I'll try and explain: Holding CAB meetings for '100% increase' on resources by way of a manual change is how we used to do it in the olden days: It's a lift and shift approach. In modern web based companies we dynamically scale the application and it's databases based on metrics (load, CPU/RAM, available workers, traffic). This way we constantly test our ability to scale and can mitigate when things go wrong eg scaling to other regions etc. Scaling should be a normal thing that happens automatically without manual intervention: My servers scale around 300% DAILY without downtime* It saves a huge amount of money not running all the machines when they are not at capacity. Everyone is happier: less engineering time for adhoc scaling, cheaper product for customers, more responsive when it's at peak usage. bromcoms takeaway is that they will carry on as planned, but make sure the resources are available with prior agreement. My takeaway is that they need to move on from the old style and that a re-architecture will save them money, improve their reliability and allow the company to scale. * Disclaimer: I work for a slightly (but not hugely) larger company that deals with web based apps, and part of my role is to scale and monitor the applciation load.

I couldn't agree with this more. I could easily manage 1000 chromebooks for the cost in time for 100 windows desktops.

And this, ladies and gentlemen, is why lift and shift is a bad approach for could services.

I very much doubt all that is user data, most of it will be transaction logs IME https://learn.microsoft.com/en-us/sql/relational-databases/logs/manage-the-size-of-the-transaction-log-file?view=sql-server-ver16

Agree. The current situation, though, is that these companies are walking all over schools.