ADMaster

My SCCM box is physical can post details of it's nic tomorrow

Hi all, I setup nic teaming on my SCCM box a few weeks ago. Server is 2016, switch is cisco 6509 I used LACP. The nic team reports functional with 2Gbps on the server SNMP tools report the po interface has 2G as well. The clients and the server are in different vlans. The edge switch is ether channel 2G fibre as well. I put the link to the test today, I imaged 2 suites at once. task manager was only report ~990 Mbps I logged in to SNMP monitor the PO interface was at 48% however the vlan interface was at 98%. I found this on cisco forums https://supportforums.cisco.com/discussion/10251226/what-max-bandwidth-l3-vlan-interface But that is not what I saw today. Any advice please?

Aside from the GPOs MS keep removing from pro, education has applocker, direct access, UE-V, credential guard, and device guard. I think applocker is key, much better then SRP. It was the main reason I moved to enterprise when we did the Windows 8.1 migration.

I use powershell. A nightly csv from my SIS, fed into a script then out comes user accounts.

Sysinternals procmon, mentioned above is the replacement for filemon and regmon which I suspect is what you were thinking of. Another tool you may find useful is regshot https://sourceforge.net/projects/regshot/

Back when we moved from 7 to 8/8.1 I put the dism command in the SCCM TS Today on W10 I only have a few machines that need 3.5 so I push it with pdq deploy with the program that needs it. I've not used MDT for a while now, but here is what I'd do. 1. evaluate if you still need 3.5, I'm guessing you have already done this since your going through the trouble to install it. 2. Create a bat file with the dism command to run as an app install. OR 3. Modify the wim to have net 3.5 enabled. The command you want is something like this. Dism /online /enable-feature /featurename:NetFX3

I +1 the deploying from source and install / customize during OSD. However to answer the UEFI / Bios question... UEFI / secure boot is required if you want to take advantage of any new security features such as credential guard. so something to think about going forward. To address the pre sysprep ghost image, have you tried clonezilla? I've not used it for about 5 years so have no idea if it will work with UEFI. Its just an idea as a stop gap measure until you sort the MDT task sequences.

dizzying, watched about half.

Once upon a time I used thinstation to pxe boot old machines to a thin client. There is light weight linux distros like puppy and chromixium worth trying too.

I'd also look at putting quotes around the paths in the robocopy command.

I've seen two threads this week that were resurrected from years ago. The new post is a new member with links to a product. I reported them both.

OK so I have two ideas regarding new / unread posts. I am typically a user of the website, but recently found the android app has a button to mark individual threads as read. 1. Can I have that button on the website too? My work flow prior to the app was to mark all as read when I was done, but it would be nice to mark it as read as I come across it. Example use cases of this would be a thread discussing a product I don't use and cannot contribute to. 2. Can we exclude certain forums from our what's new link? There are some sections of the forum dedicated to a product I don't have, or region I'm not in. I accidentally replied to a post in one of the gfl sections. When on the site and I see a lot from one forum I just go and mark that forum as read, and on the app the button is just as fast. It would kind of be an inverse of the subscribed forums. Two ideas for implementing this; Create the option to have exclusions as above Create a button for What's new in my subscriptions. named something shorter of course.

Do you have the install applications step configured to continue on error? you should have got some sort of on screen error of the failure otherwise. Do the applications deploy outside of a TS? Create a user collection with just yourself, deploy an application as available and run it from software center. Another difference between applications and packages is you need to configure the application with success conditions. Look for this product code, file or reg key etc. If you don't have that set the application will always report as failed even if it does install.

Another bit on applications, you can setup the install as an application and point it to the script if needed. This will sort of give you the benefit of both methods. Most applications I use the native installer and call the appropriate switches. However there have been a few apps that need a bit more tweaking, so I point it at an install script.

I remove apps during OSD and do not modify the win. It works great for me, I do all customization on the fly during OSD. No messing around with the wim, capture task etc. Here is the script I use to remove apps. I copied / cobbled it together back on windows 8 and just added to the $appslist for windows 10. You should just need to change the $applist to suit your preferences. remove-selectedw10.ps1

No they just need to be there prior to that user opening chrome for the first time. I don't use mandatory profiles so that could be where your getting hung up. What happens if you add these files to the mandatory profile? The only issue I see with that is the preferences file is modified and added to after chrome is launched. homepage, installed extensions etc. If there is already a preferences file in the mandatory profile maybe it just needs to be modified. Just for your own testing / sanity check either create a new local user, or domain user without the mandatory profile. EDIT: to speed up testing of changes delete this folder C:\Users\username\AppData\Local\Google\Chrome Then add back in the first run and preferences files.

Does this bug effect only the DE / international version? I just checked on a 1607 PC and all looks normal to me, I'm using en_us though. Thanks,

The first issue sounds like the trouble I previously had with classic shell. Another quick fix to that one is to kill the exploer.exe and re run it, not the most intuitive for an end user. I ended up removing classic shell from my build. It worked great for W8 but not W10.

Have you tested this with a new user on the PC? my scripts are meant to be used during OS deployment. If you are doing this after a user has logged on for the first time you will need to add in some sort of loop to add the files to their profiles. I just imaged a few machines today and used chrome 59.0.3071.115 enterprise x64. The settings applied as expected, no welcome screen. I have just updated my app in SCCM and am testing with V60 now.

Check / try a different ethernet cable / socket. Try again once the other deployments are finished or farther along in the process. Two years ago or so I had a PC that just would not image, identical to the one next to it. Booting into windows it showed a connection, same driver etc. The only difference I saw was it was connecting at 100 bps vs 1 gps. I moved it to another socket and it it imaged fine. This year I imaged a lab of older PCs getting them all going at once wasn't going to happen. I checked bandwidth and it all looked OK, but some would not image until others were finished.

Vlan 120 client ip and gateway are in different subnets Mask needs to be 255.255.248.0 or with the current mask bring the gateway to 10.16.115.254 The gateway of your vlan1 is in vlan 120 ip space You need to sort out your overlapping addresses Use an online subnet calculator with cidr your looking at /22 and /21 right now Your /21 can be made into two /22s Also there doesn't appear to be a method to your gateways Use .1 or. 254 and stick to it .55 is odd to me. It makes documentation easier later on I also number the vlan to the third octet where possible Your vlan 120 should start out at 10.16.120.0 or make it vlan 112 It just helps to keep it all straight in my opinion Edit I see your switch config and dhcp scope list vlan 120 as 10.16.120 and your most recent post lists it as 10.16.112 I'll just assume typo and forget most of my post

To the Robocop error 1 run as admin 2 exclude junctions Iirc I had to do this because the system somehow had a loop of junctions Just allowing the system to follow them made an ever growing path.

Aliens/ UFO beaming something up. Wish I had got the photo a few minutes sooner the sun rays were a bit more visible. I will try again tomorrow.

Since v57 or 58 you have needed to make and copy a prefs file. See my other post /showthread.php?p=1572171

I do not deploy anything on mass to the users, but I use 7-zip and deploy it to the occasional power user if needed.