Back up servers IT support :-)

Some how I’ve ended up looking after IT in the smallish firm I work for – circa 50staff….

I learned how to code in BBC basic, the ZX Spectrum and Fortran 90, since about 1997 I’ve become a microsoft slave and not really kept pace with IT…

We have three offices spread across the north west and our main server is located in one of these offices.  The internet on the business park keeps dropping and it’s playing havoc with us accessing databases and vpn etc.

Now we have servers in each offices, they have lights that flash and fans that whirr so I know they do something, mainly local office project information.

I want to spread the function of the main server across the three offices to provide some resilience to our systems and I know our IT supplier will bamboozle me with buzz words and reasons why not.

Can you lot point me in the direction of some good places to read up on this stuff so I don’t just come across as I want but have no idea how?   It all seems pretty simple but I know we’ve been discussing this over the years and it hasn’t been resolved…

Thanks in advance

Steve

Steve no disrespect to you and your enthusiasm, but a firm of 50 ppl with three sites should either have a proper IT support team or use an external support service. This sounds ripe for disaster, and you will be at the forefront of the blame.

Otherwise dump the physical server, use a cloud based service, and (or just) get better internet connections to the sites.

Was going to. Suggest getting it all hosted somewhere like rackspace, ukfast etc. And maybe use your existing set up as a fail safe /backup that you can switch to?

We moved from in-office server and a ‘unique’ (bodged) system to a proper IT company on Office/MS/Sharepoint etc in 2019.
It has been the best move ever, upped productivity, is stable, maintained and is an IT company that are very responsive when some issues raise.
In these days of increased cyber threat and how vital IT is to any company, get the professionals in.

If you are sharing databases and you are having connectivity issues then prepare for a world of pain.  Yes, it is quite possible to do and if you you are using SQL server (not Access) as the database you can make it work but if it isn’t designed to work that way then it’s going to be painful.  As MOAB did, I moved our linked servers to Office365 in the cloud and life is much much better.  This will depend on whether you can do that with your databases.  This is going to be a big project and don’t just do the easy bits and hope the rest works, you need to know the whole picture before you start.

Sorry, I’m afraid I can’t recommend any reading though :(.  It really depends on exactly what you have at the moment as well as how you authenticate to your servers.  It’s a big topic

as above, we binned off all of our servers, wasn’t worth the trouble maintaining them in our head office and trying to keep the other sites as satellites. Exchange was the first to go, and didn’t take long before we moved everything else.

Prior to that we had at least one server in every office internationally, and being as there was 2 of us.. not international.. was not very easy for us to look after.

If you are a small team, then leveraging someone else to burden the hardware is a no brainer.

Rather than reading, try something like Pluralsight, online training company its about 400 for a years access to their whole library of training, if the company is paying for it no probs, if you are paying yourself then they frequently have cut price offers.

I am sure there are other companies with similar offerings, but I have used Pluralsight and found them pretty decent, it can be a little difficult finding the right course, the listings aren’t well organised but the actual courses are quite good.

And yeah move to the cloud man…..

Just remembered, if you really want to read something what I did was buy a huge honking book on whatever flavour of Windows server we had at the time and read most of it so I knew what I was doing.  Nowadays I would recommend focusing on writing down everything you expect your system to do after your IT supplier had worked on it e.g. do you expect the databases to sync across offices, do you expect the same documents to be available across offices, do you want to be able to use the same login in each office etc.  Don’t imagine that anything is obvious otherwise you won’t get it (I say as someone who is into year 2 of a 4 month migration as the person in charge didn’t write it all down so got what the suppliers could easily do).

if the internet keeps dropping, any synchronisation to the other offices is just going to be another thing that fails. I suspect Office365 and shared documents via OneDrive or similar would meet your needs (or google docs if you’re not tied to Microsoft) but that is just an educated guess without further info.
Stable Internet is a fundamental service these days, so I’d look to fix that before trying anything more complicated

You also sound like the ideal customer for AWS or Azure Cloud, where you offload the server maintenance to the cloud provider and you manage the user admin – this is a very simplified view as you’d need to investigate this properly before going down this route. I wouldn’t bother with rackspace or similar these days as they are a dying breed, they just can’t compete with the big three cloud providers (Azure, AWS and Google) who can provide traditional “server” offerings or more cloud-native setups.

There are a number of IT consultancies that would help you navigate this, a good starting point is to look into MeetUp and find the local Azure or AWS User Group in the area.

Drop me a DM when you get Ransomwared.

More seriously on premise data is like crack cocaine for the script kiddies. Hope you have MFA deployed,  mobile devices encrypted and lots of locks and bars on the windows and a ruthless long password policy in place.

M365, get it all in the cloud.

We moved our servers into hosted servers (Azure) and O365. Never looked back.

I got a company to do it all and it was relatively easy for them too. Can pass you their details if you like….DM me.

We’re on 365, using SharePoint is simples. Some external it company maintains it and we just use it. We’re(largely)  all home based and can collaborate easily.

The first thing I’d do is audit what you have. You cannot improve matters until you progress past “it has lights on it.” You need a full understanding of the current setup, what systems are where and how they talk with everything else. Document it all, then work on a Technical Improvement Plan.

If you fall at this first hurdle, say “no” and kick it back.

^Very good advice.

How can you understand what you want, if you don’t already understand what you have, etc…

Right you lot, you’ve done me proud, I’m clearly out of my depth and you’ve given me a few icebergs to hang on to whilst I go back to our service provider. I will go over the thread in detail tomorrow and no doubt ask more stupid questions!

Honestly, you need to FULLY understand what you’ve got before you decide what to do.

With respect, it sounds like you are a very long way off a good understanding at least.

The fact that the organisation has put you in this position, is *not* a good sign at all.

If it is REALLY* just office apps and related docs then get it on 365. Service provider may not be helpful (obvs)

* It very much might not be this.

Raise concerns to senior leadership asap, especially if any of this stuff is critical to the organisation functioning

You aren’t going to go from 0 to  the years of experience to know what’s what by asking questions on STW. Sorry.

Get professionals in who are independent from the service provider.

Good luck

Lots of good advice on this thread already.

The core of your issue is you have on-prem IT that’s accessed by users from remote locations and your Internet is garbage, that’s about the clearest case imaginable for migrating it to the cloud or, if your MSP offers it, having them host it out of a data centre but that will likely work out more expensive and will be a real pain if you ever want to move away from using them. I wouldn’t just go out and buy some AWS/Azure credit and stand up a server – work with your MSP, they likely already have a managed cloud service (unless your MSP is Dave working out of his garage). If you really want to get more hands on then discuss what access you’ll have (to both the cloud provider portal and your company’s servers that are hosted there), do not expect admin access.

It’s very easy to deploy in & connect to the cloud in a way that will give you eye-watering monthly bills but one of the big advantages of cloud is you can size up and down quickly so your MSP should be advising ways to start small but also that you need to bear in mind costs could quickly rise depending on the demands your company is making.

It sounds like you also need to improve the Internet connection at at least one of your sites, who is your ISP and telecoms provider currently? Is it your own building where you can install your own links etc.? Lots of comms companies will be able to provide redundant links but it can get very expensive (especially if it involves ground works to route it entirely separately from your existing links), you probably need to do some analysis of the cost of lost productivity for sites when their Internet goes out. It’s also something to bear in mind if you want to migrate more and more of your internal IT to the cloud in future, it’s quite easy to get to the point some/all of your users can’t do anything if the office Internet is down.

Lots of good advice above and I hope you realise the risks involved in current situation and also amount of planning and budget you need to improve the situation. It may well be that company does not want to spend enough to improve situation until total system breakdown.
And it is worth noting that going to cloud does not always mean reduced costs.

(I say as someone who is into year 2 of a 4 month migration as the person in charge didn’t write it all down so got what the suppliers could easily do).

Ha! One of my ex employers did a migration from an antiquated handbuilt/custom/internally managed system to a fully supported supplied system and we wrote down *everything*. In detail.

The supplier quoted against this list and 6 months into the pilot program it became fairly obvious they hadn’t read (or understood) the brief.

Ended up taking them 3 years and a custom build of the software to meet even the absolute bare minimum requirements to make the system vaguely compliant. And another 2 or 3 to meet their contractual obligations. That they said they’d be able to fulfil in about 18 months. From talking to colleagues they were still finding (and fixing) basic operability issues at around the decade mark…

Cloud is your answer, maybe not your budget

Ha! One of my ex employers did a migration from an antiquated handbuilt/custom/internally managed system to a fully supported supplied system and we wrote down *everything*. In detail.

The supplier quoted against this list and 6 months into the pilot program it became fairly obvious they hadn’t read (or understood) the brief.

Ended up taking them 3 years and a custom build of the software to meet even the absolute bare minimum requirements to make the system vaguely compliant. And another 2 or 3 to meet their contractual obligations. That they said they’d be able to fulfil in about 18 months. From talking to colleagues they were still finding (and fixing) basic operability issues at around the decade mark…

Moving to the processes supported by standardised software is usually easier & cheaper (in the long term)…

Im sorry to be negative about this (and its not reflection on you) but I do have some experience in this area. Your organisation needs to take a more responsible and strategic approach to this.

Moving to the processes supported by standardised software is usually easier & cheaper (in the long term)…

Not if it means you lose a load of accreditations, and have to stop selling stuff in multiple markets, and multiple segments of other markets… there were many many times the supplier said “why do you do this, no one else does” to be met with a massive government/military/governing authority document and be told “To fulfil this”. Even without those legal bits n pieces, we’d have lost (quite literally) millions of pounds a month due to their inability to handle certain relatively simple aspects of the process flow.

AFAIK the later versions of their software have all this functionality built in. So they did learn.

You already have an IT supplier? What are they doing for their money?

Your idea of spreading ‘ the function of the main server across the three offices to provide some resilience to our systems’ seems like an odd solution to the problem of bad connectivity. That connectivity should be your first problem to address. Once that is done then you can review your resilience and recovery plan which should already be well-documented since it sounds like your business depends somewhat on its IT capabilities.

A book to read? I highly recommend ‘Crash: learning from the world’s worst IT disasters’. This is an ancient book now, but provides some helpful insights into how things go wrong – hint, planning is a big factor. Some of the disasters seem fairly low key compared with the recent Crowdstrike SNAFU and other high-connectivity freakouts.

And it is worth noting that going to cloud does not always mean reduced costs.

As one senior director of IS described to me BITD ‘once it’s in the cloud it isn’t coming back out [to physical servers we would have to get]’. These ongoing costs will need to be assessed against the costs of your current systems and appropriately modelled for the risk costs to help understand the ‘real’ cost.

‘the cloud’ only means that your data and some computing is done on someone else’s machines. There may be information policies that will mean geographical constraints to where ‘the cloud’ can be for you. If your Internet connection is variable then a move to the cloud may not be as smooth for you as for folks with great connections.

An ill-conceived move to the cloud may end up being expensive, difficult, and certainly risky. I think most folks here seem to be saying things like:

• document your current system and processes – this is already in your SOPs and operating manuals right?
• assess you current situation
• determine what your system needs to do now
• determine what planned changes (improvements) are needed in the next 1-5 years (some of these may be simplification of software choices – swapping to standard things from custom things)
• and so on

As an aside,

Even if you ultimately decide to change nothing and/or outsource the IT piece, the time to document everything is when you aren’t in the middle of dealing with an outage.

The thing that stood out as a big red flag for me was this:

“I know our IT supplier will bamboozle me with buzz words and reasons why not.”

Sounds like you’re big enough to need an IT team and small enough not to have your own IT dept, so an MSP is normal in this situation. However, they should be providing a service that meets your needs, not what is best for them. If they can’t talk to you in a way that is easily understandable then it sounds to me like you should be shopping around for a new provider that can help you make the best decisions and not giving you reasons why things can’t be done. Regardless of the tech involved you need someone who will be proactive and not dismissive.

Techies that know their stuff but can talk to you without buzz words and jargon do exist, they’re just hard to come by…

As another aside,

I used to work for a MSP. If you’re looking at a cloud migration, depending on your industry this may raise compliance issues. For instance, we had customers whose data had to be stored on servers wholly contained within the UK in order to meet various regulatory demands.

We were our own cloud provider, our servers were in a couple of datacentres in London so for us that was fine. But we had to be just a little bit careful when offering services like M365, SharePoint, OneDrive… Yeah, your data is in Telehouse North but your email is in California.

GDPR is certainly worth bearing in mind but cloud providers have been catering for this for years now and have configurations that mean data can stay in the UK or Europe.

We were our own cloud provider, our servers were in a couple of datacentres in London so for us that was fine

That sounds more like an MSP with a couple of data centre locations providing a hosting service, not a cloud provider…

:shrug: Marketing innit.

Not if it means you lose a load of accreditations, and have to stop selling stuff in multiple markets, and multiple segments of other markets… there were many many times the supplier said “why do you do this, no one else does” to be met with a massive government/military/governing authority document and be told “To fulfil this”. Even without those legal bits n pieces, we’d have lost (quite literally) millions of pounds a month due to their inability to handle certain relatively simple aspects of the process flow.

AFAIK the later versions of their software have all this functionality built in. So they did learn.

Thanks son…

I’m in my 40th year in the ‘business’ and well aware of the pitfalls, with the scars & bonus cheques to prove it.

Data residency of cloud services is easily solved, the major providers allow you to choose your location/region and UK only would be available for pretty much all the services named in this thread.

I work with heavily regulated, critical national infrastructure customers who host significant amounts of services in public cloud, with UK residency. Public cloud is also used by NHS, MOD, HMRC etc. (all public references)

There may be niche or very new services that take time to arrive in certain regions but I doubt those are anything you’ll be needing.

Of course, if you want to have your M365 in California you still can.