Geoff

I moved from SCCM 2012 on 2k8r2 to SCCM CB on 2016. I would upgrade SCCM on the 2k8 box as far as you can (2012 R2 v1607 IIRC) then either do a site migration or a site backup and restore to the new VM.

Sounds like your installer is guffed if it works manually.

It entirely depends on if you remember to go back into the application deployment after setting it up and unticking the box that wasn't there before. But you can only do that if you set things up in the right order (create new app, deploy, alter hidden tick boxes, create supersendence rule). https://foxdeploy.com/2016/01/21/sccm-controlling-application-supersedence/ Here is the uservoice entry if you want it fixed. https://configurationmanager.uservoice.com/forums/300492-ideas/suggestions/11532669-fix-supercedence-behavior

I use supercedence here, so for example I have Adobe Flash doing this. I have a custom PSADT script that handles the install / uninstall process for me so I do not need to tick the 'uninstall' option. It will just install over the top of the old version. If you did have something (say an MSI installer) that had to be removed before you could put the new one on, you would tick the box. One gotcha that might trip you up is that clients will immediately jump on the new version once it's been distributed rather than paying any attention to what your deployments say.

I believe outsourcing your DPA / GDPR officer is a possible solution if you can't dedicate a member of staff to it. It is something my employer offers as a service.

This table has the same mod kit as mine. Although I replaced the original scoreboard LCD with a colour one too.

Not exactly the same, but similar. I bought and restored a Midway Theatre of Magic Pinball table. Internet Pinball Machine Database: Midway 'Theatre of Magic' As the project you are considering is going to be a lot easier than mine (new components vs 90's tech) I think you shouldn't have any major headaches. My only advice is to check your electrics. Make sure all the buttons / lights / etc are running at the right voltages. I had great fun replacing the original sound gear in mine with a new uprated board. I also had a hell of a time replacing the incandescent bulbs with leds. I wish you good luck, the end result is worth it.

https://www.amazon.com/HP-Procurve-E2620-24-J9623A-ABA/dp/B005XUE486

Ok this assumes that: VLANs have a one to one mapping to an IP Range. You are using the Procurve as a router (you have ip routing in your config). You have WoL set to 'Subnet directed Broadcasts' in your SCCM Site Config. In the SCCM port settings, your UDP WoL port is 9 (the default). On the 'sending' VLAN (Where your SCCM Site Server or Distribution Point is): ip udp-bcast-forward ip forward-protocol udp 9 On the receiving VLAN (where your clients are) ip directed-broadcast You can verify the packets are working using wireshark. The capture filter you need is 'ether proto 0x0842 or udp port 9'.

Yes I have this working here. There's three bits you need to get right. (Assuming you have VLANs) Allow directed broadcasts from your source VLAN to your destination VLAN in your network switch(s) config. I can provide an example for Procurve. Your PCs need to support WoL properly. This is simply a case of getting the BIOS settings right. I have Dell Optiplex 760 workstations so I can only give these as an example. Your Windows NIC drivers also need to support WoL and have the options set correctly. Again anything with an Intel card is ok (like my Dells). Realteks/Broadcom, not so much.

Yes, you need to do this in each server that is presenting the file share by the common DFS namespace / share. The simplest way to make this consistent across multiple servers (especially if you have a lot) is to script FSRM (Powershell in newer Server versions) to do the quota setup. That way you know the configuration is identical. There's plenty of scripts floating about that let you backup / restore FSRM quota settings. e.g. https://github.com/monosoul/fsrm-quotas-backup

We have double uplink ports anyway for speed and resilience. VLAN 1 is tagged over the main uplinks and untagged via the OOB port. The OOB port and the Serial KVM connect to a separate switch to the core.

Dedicated Ethernet port which goes into the local network stack. Serial Port is also plugged into an Serial KVM with a port on the Management Net. Core network stack is a HA pair of switches with dual uplinks to each edge switch. Uplinks from site <-> site is via MPLS with a backup IP VPN link. So yes, fully isolated and redundant connectivity. It's impossible to lock yourself out of a switch remotely.

You can use ACLs if your switches support them. However my environment is complex so we used a firewall for performance and manageability reasons.

For switches we have an out of band (where possible) Management VLAN. This means the switches aren't addressable on the 'production' VLAN(s). It also means you can mess up a remote switch config and still access it. We also segmented our network. So we have 'workstation' VLANs and 'Servers' VLANs. There is a firewall (pfSense VM) sitting between these VLANs that only allows the traffic through we deem acceptable.

https://support.office.com/en-us/article/After-installing-KB-3213656-or-KB-4011039-merged-table-cells-don-t-work-correctly-in-Word-or-Outlook-8c7af9eb-9e48-4e1e-8c13-6340ede4acdc Fix in the October Patches

Generic Powershell script I have in SCCM for stuff like this (*waves fist at Microsoft*) $SearchUpdates = dism /online /get-packages | findstr "Somestring" $updates = $SearchUpdates.replace("Package Identity : ", "") | findstr "KBXXXXXX" DISM.exe /Online /Remove-Package /PackageName:$updates /quiet /norestart Replace the "Somestring" with something unique in the update title, likely the KB or MS Security number. Only used this on Win10 as well.

Even when I uninstalled WSUS on our branch servers and ticked the box that said 'remove content' it still left files strewn everywhere.

Consistency if you have less than 16 sites across a WAN. I imagine you don't.

Dilbert Daily Strip

We have VSAN, running on a mix of self built Supermicro kit and bespoke purchased Dell servers. Our DR is Azure Site Recovery.

I have a powershell script that adds printers based on a database. Users self manage their printers via a web page. Alternatively there is a Powershell GUI app they can run to pointy clicky things. I have found you have to wait around 20 - 30s after logging in on our hardware for printer mapping to work. This is due to using rubbish printer drivers that aren't natively written for Win10. We've standardised on about 3 drivers now and fixed the issue.

Yep, that did the trick. TY

I goofed this morning (skipped my coffee) and then compounded the error by deleting the 'Site status for the hierarchy ' report instead of the subscription. How can I get it back?

Ok, I'll be specific. A single application with a single 'Script' delpoyment type that is deployed to a collection called 'Windows 10' with a deadline of next week. The same is also deployed to a collection called 'Test' with a deadline of 5pm. Which deadline will apply?