Michael

The vast majority of G Suite is down port 443, so I wonder if it's something like SSL inspection which is causing the lag? You should be able to add exceptions.

If you open a Security group in AD, on the General tab, have you populated the email field? Here's how to make your Exchange server work with G Suite. You do need to leave Gmail enabled, as Classroom uses functions/GUI of Gmail but will look to the DNS MX rule. Even if someone opened up the Gmail app and sent an email, any replies would end up on your Exchange server. I've written about GCDS + GSPS vs AAD. I've also linked to the YouTube clip I used for reference.

I'd have thought so. I am presuming they're hot plug, but you're just as well to shutdown the server, remove them and install your new 2TB drives and build the array. As you say, if you build a new array using your 1TB drives, these would be erased anyway.

SATA drives can be connected into SAS controllers. SAS drives cannot be connected into SATA controllers.

I've never used this function. I suspect it's probably writing lots of 0's and 1's, hence taking so long. I didn't think you had any data on the array as yet, only one massive volume?

I'm not sure what you mean by setting the drives to be erased? Destroying the array will destroy any data.

Hi all, I need to clarify some steps to do with an O365 Migration as follows - Source is an LA based O365 solution, so I don't have admin access Destination is a vanilla O365 solution, so I do have admin access 1. Setup new tenancy and provision users with a CSV 2. Ask LA to rename nominated accounts to onmicrosoft.com, release .sch.uk domain, update TAG and set passwords to a complex one of my choice 3. Update nameservers at new registrar 4. Complete DNS changes so .sch.uk is registered in new vanilla O365 tenancy 5. Any new emails received will automatically arrive in vanilla O365 tenancy 6. Setup MS Outlook and connect to required onmicrosoft.com account in turn - download mailbox, contacts, calendar as required 7. Merge PST in MS Outlook to new vanilla O365 mailbox in turn 8. Check for any OneDrive files (there are no Sharepoint sites), but I'm expecting few OneDrive files/folders for most users 9. Once complete configure AAD for SSO logon 10. Drink lots of coffee Does this sound about right in terms of strategy?

I'd have to agree with the above - you're essentially talking about 21,000 odd SQL databases, training and managing the remaining third parties still required. It's definitely not a small task, but then lets be realistic, schools are going to make decisions at different times/different years. I can only speculate whether any of the competition have even contemplated having many thousands of schools to support with MIS issues, on an on-going basis.

You're right I could, but that's on the presumption that everyone's using scripts to create users in AD. Some do and some don't.

I can't really comment on SIMS Primary as I haven't seen it, but as others have commented, SIMS 7 does need a full re-write. It's perfectly possible, but the logistics and man hours required to do so could be very costly. I suppose other considerations for a potential buyer are - - Existing customer numbers - SIMS still has a strong customer base, but the numbers are dwindling overall year after year. - The competition - there's plenty out there already with growing customer numbers. The landscape looks very different to 15 years ago. - MATs - they change the dynamic of how schools work. All legacy MIS solutions are very much standalone and that doesn't really work for MATs. - Procuring - clusters of schools or MATs like to procure a single solution that'll work for all. In the context of SIMS 8, there is no SIMS Secondary at the current time. As to what exactly the differences would be between SIMS Primary and SIMS Secondary, given SIMS 7 does both is unclear. - It's not just about re-writing existing code. It's also about writing new modules altogether - good examples are MAT based dashboards, native cashless solutions and safeguarding modules. SIMS 7 doesn't have any of these functions but as above, to the competition they're nothing new either. - Hosting such a service can be tricky, but with the likes of MS Azure, Google Cloud and Amazon, the structures are already built. It's all about finding that right balance.

Hi all, I've had the opportunity to setup GCDS (Google Cloud Directory Sync) + GSPS (G Suite Password Sync) and thought I'd compare it to Azure AD Connect (AAD) which I've used for years. Typically I'd add G Suite as an Enterprise app in Azure and make everything 'talk' this way, but getting AD to sync with G Suite is looking at things from a different perspective. GCDS is definitely more complicated, with a requirement to understand LDAP and base-dn's which make my head hurt after a while. Why Google haven't provided a tick box GUI for this excercise (just as AAD has), it'd make things easier for large structures. Thankfully hopping onto , there are some good tutorials on there, but some of the menus have changed. Goes to show how much development Google have put into it! Once all configured, unlike AAD I had to create a script along with a Scheduled Task to allow it to run every 30 mins. Script "C:\Program Files\Google Cloud Directory Sync\sync-cmd" Arguments -a -c "C:\GCDS Settings\GCDS-Settings.xml" Again unlike AAD, GCDS doesn't handle passwords as this is done with GSPS. Thankfully I didn't have to do too much with this, other than authorise the tenancy and was straight forward, however... If you create a user in AD and they sign into Windows the first time before the account is visible in G Suite (think about the GCDS script every 30 mins), they'll be forced to change their password a second time before they can use their G Suite account. I also found all existing users in AD need to change their password (it could be kept the same if done by an admin), as this generates a token, which GSPS then reads. GSPS synchronisations were generally pretty quick. I then continued to dive in deeper and start experimenting with disabling users in AD, but I very quickly started getting errors. I then found on the various menus there's an Exclusions tab which was set to 5%. This controls the maximum number of possible changes per cycle. Not ideal if you're only experimenting with a handful of users, so I increased it to the maximum of 80%. I also found the only reliable way to disable users, was to disable them in AD and then move them to a seperate root OU in AD, rather than say a sub-OU where your staff objects are for example. Not a major issue and I'm sure it's possible, but I couldn't figure it out. Increasing the number of changes per cycle from 5% to 80% doesn't concern me, as GCDS has a Simulate Sync option which was really useful. This also provides logs, so you can fix issues before they happen, unlike AAD where you find out afterwards if there's an issue. So I hope this helps someone else out - I know not the most thorough of reviews, but having done both I now know the pros and cons of each! AAD Pros - probably easier to use/setup than GCDS Pros - user and password sync in single solution Pros - more GUI options to assist with setup Cons - definitely heavier on memory resources Cons - no simulation sync, fix issues on live system GCDS Pros - light on memory resources Pros - useful log information and liked the simulation sync Cons - definitely harder to setup, but good tutorials online Cons - won't schedule itself to run every 30 mins Cons - GSPS syncs passwords separately. Not sure if this is a good or bad thing?

It never occurred to me to look at the requirements. File Stream could previously work on Windows Server, but it was the installer that prevented so making it a faff. Even so I think I'd still run it from a Windows 10 VM, so it leaves your server free to process other services.

Best advice from previous experience is - Configure a VM running Windows 10 and install File Stream. This'll be your mechanism to migrate data in chunks, depending on your connectivity speeds as File Stream officially cannot be installed on Windows Server. Migrate shares 'as is' and thereafter split them up or create new shared areas as required.

Sometimes overlooked, but make sure the account password hasn't expired. If using AD Connect, specify the account not to expire in AD.

Famous last words 10 years from now 12TB won't necessarily be considered massive anymore!

So overall you'd have 12.7TB usable on RAID10 thereafter, but you can then split these into volumes within the OS.

Not all GPOs apply to Pro, only Ent or Edu versions.

I deploy LTSB16 & LTSC19, both can be activated with MAKs.

If you run your drives through a RAID calculator (they're really useful) it comes out as follows - 12 x 1TB in RAID10 = 5.5TB of usable space. 12 x 2TB in RAID10 = 10.9TB of usable space. This essentially means you're losing 5.5TB of space and the RAID array won't see the full 2TB due to 1TB drives being part of the same array. Unless HP have somehow introduced a new RAID model, it's always been the case that the smallest drive capacity takes priority.

I suspect the 2 x 600GB were meant for the OS which is OK too. Not quite sure why you have 6 x 1TB and 6 x 2TB drives. If you require one huge RAID10 volume, you'd be better off with 12 x 1TB or 2TB for example. You no doubt already know, but regardless of RAID type, drives need to be the same capacity/speed. With RAID6 and RAID10 you can add more drives as required, but with this configuration you can only have - 1 x RAID1 (600GB's) 1 x RAID10 (1TB's) 1 x RAID10 (2TB's)

Re-reading your post, if you have 4 x available slots, I'd buy 2 x smaller capacity drives - say 300GB SAS for example. Configure this as RAID1. Then configure the remaining/existing drives as RAID10 or RAID6 as required. Then as I wrote above, install the OS on RAID1 using Assisted Setup and configure your volumes within the OS thereafter on the RAID10 or RAID6 array.

So from experience the last HP Gen 10 server I configured, I set the arrays to RAID1 and RAID6. Using the Assisted Setup I then select the RAID1 volume and install the OS as normal. As far as I could make out, there was no facility to specify partition sizes, but this may have now changed with later/newer firmware. Upon entering the OS, you can then create the required volumes by splitting the RAID6 array up, or as one large volume as required. Windows Server 2019 doesn't require terabytes of space to install. Probably the absolute minimum would be 100GB. Same with Windows Server 2016.

I suspect this has something to do with RAID arrays and assisted setup, for example on Dell and HP servers. For example I'd expect RAID1 for the OS and RAID6 or better for data volumes.

Unless you cut the Ethernet and pull it through with fibre attached. It sounds like it's been pulled through MDPE which is typically 25mm thick in most cases.

I guess if there's power going to the shed, there must be a duct. You could run fibre in it and it wouldn't suffer from interference, but logistically this could get complicated and expensive. I suppose even if fibre is route a long way around, it doesn't matter. Even OM1 will do 1Gb/s up to 275M. I'm guessing the existing Ethernet cable is running on a catenary wire?