YPCC

try the standalone install for that KB? https://support.microsoft.com/en-us/hotfix/kbhotfix?kbnum=3122637&kbln=en-us

i had this issue, seemed a little bug where even though i specified "Use New Software Center", it was still using the old software center, no matter what i did. Turns out some people have had to manually force the new software center to be used by deleting the old shortcut and placing a new shortcut which point to the new software center .exe file (its under the ccm folder, clientux or something like that).

interesting one. Wasnt aware 1511 required 4.5.2. May need to review my migration plan. In your case, from what i understand this is the series of events taking place when the client installs. - run ccmsetup from your location will sitecode etc paramters - Its then querying AD for a Management Point (Performing AD query: '(&(ObjectCategory=mSSMSManagementPoint)(mSSMSDefaultMP=TRUE)(mSSMSSiteCode=ABC))') - Off the back of this query, it is automatically finding a suitable distribution point (Found local location 'http://SCPRIMARY.corporate.abc.com/SMS_DP_SMSPKG$/ABC00003') - then downloading the ccmsetup.cab (GET 'http://SCPRIMARY.corporate.abc.com/SMS_DP_SMSPKG$/ABC00003/ccmsetup.cab') - which contains the manifest (Successfully extracted manifest file C:\Windows\ccmsetup\ccmsetup.xml from file C:\Windows\ccmsetup\ccmsetup.cab.) Try this: - Make sure all the files required to install the client are on the machine then take the machine off the network and install the client. This way it cannot connect to AD, therefore wont find the management point and distribution point. If that method works you will need to consider how to arrange offline installs. - I've never really looked into how distribution points store content, but if its stored in its original format, maybe you can consider editing the CAB file on the DP by importing the manifest XML file into it. Then even if it downloads the cab, it will contain the manifest you want. Just a thought.

are you patching any other users in the organisation? If yes: - Patch those first, phased approach. Then leave these business critical users till the end of the patching cycle. If no, then a choice of options: - attempt to determine "less" critical users/pilot group from those in the call center and patch those first, followed by rest a few days later - make patches available with a 7 day deadline, so users who want to patch can do so at their convenience. Others get patched at deadline (this empowers you users but management may not like this) - Break down the 1000, into 10, 30, 200, then rest over a 4 day period Lots of choices, management will just want the assurance that you're don't take down the whole lot with a bad patch. Basics of any deployment (especially patch management) should be using a phased approach. Start with a pilot, and slowly ramp up the numbers.

+1 for side by side. As you already have 2012, a forgotten bonus is the powershell ability. Im planning on going from 2012 to crrent branch soon and will be using powershell where possible to import and export items such as applications, task seq etc. Need to create 100+ collections? No problems, powershell it Sorry for going somewhat off topic

Sccm is designed to manage 500 devices or 500,000 devices, with no additional requirements apart from extra resource (ram, hdd, cpu etc) The point im making is unfortantely its not the greatest tool when it comes to deploying something immediately (as in under 30seconds) Once you deploy as required and set to install as soon as possible, the machine must contact sccm to see that a package is available. Until it makes contact, it wont know theres a package available. Easoest way to get around this, install "right click tools" and run a machine policy eval over the collection.

How good are you with SQL? Access the database via Sql management studio, open the gs_computersystem view, link it to gs_system (resourceid) and see what the sccm Database is storing for models. Or just right click a machine in sccm > resource explorer > computer system. Does it show correct model? Perhaps SCSM is storing a cache of data thats no longer up to date? No idea how scsm works but sccm should have correct data. If it doesnt you have an issue (a big one if you use query based deployment)

Hi all, Looking for some tips on this one as its not the most straightforward of activities. Scenario: We are planning to outsource our imaging to Dell, essentially this means we will create a image, and Dell will preload it onto newly purchased hardware. Therefore when the end-user gets their new laptop delivered, they simply connect the network cable, switch it on and follow the very basic on screen instructions (choose country etc), and away they go. Based on the country the users choose, lang packs are installed, time zone set, locale, and even the computer name is prefixed with the country code. We're familiar with this part. I'm curious as to whether anyone has been able to implement a recovery partition using SCCM task sequences. A scenario where this would be required is when the end-user has issues on their machine but no local IT support is available to rebuild it. We could instruct the end-user to restart the computer press a "F" key and recover the machine back to its original state. Thus totally eliminate the need for local IT which can be VERY costly in remote areas/countries. I know both pre-staging and recovery can be achieve as separate functions. Its combining them into a single, automated, deployable solution that is my task. So curious as to whether anyone has done this before and has any tips that could speed things up on my end.

2 options: 1 - Set a maintenance window on your collection so that way nothing can occur on the server unless you drop the main windows for the duration of the patching. 2 - Create a new "Client Setting", increase the software update cycle frequency so it checks for every 15mins as you wanted. Deploy this setting to your servers collection. Be sure to raise the priority of this setting so it takes precedence over your default client settings I always use maint windows on our servers, even if to prevent a accidental deployment. I have a powershell script that then deploys my updates for me and creates maintenance windows so takes away all the manual work from me

your querying "installed executable". I usually opt for "installed software" as in our environment that seems to be the most reliable method.

I also agree that uninstalling anything with "Java%" in its name is dangerous. I prefer instead to run a query that brings back the Java products installed as well as the uninstall string for them. I then create a script that contains all these uninstall strings. A tad bit extra work but safer than removing all Java items on your machine.

Using the application model to deploy scripts requires just a little more caution. - Double check your detection method is correct and looking for the exe in the right location (prog file x86 / prog file etc) - Double check the deployment type I actually "script" and not "Installer". I've made this mistake before, (https://technet.microsoft.com/gg675930.Fig_2_Each_application_can_have_multiple_types_of_deployment(en-us,MSDN.10).jpg) - Does anything actually get downloaded into the "ccmcache" folder when you press install in Software Center? - Whats the error code being displayed in Software Center? Press install, wait for the failure message and click "more information". (https://social.technet.microsoft.com/Forums/getfile/165322) Also you could simplify your script if preferred, you could have just 4 lines: @echo off cd %~dp0 (this isn't required, but just makes sense to have if you will be manually installing on some machines) setup.exe /s exit

Open your task sequence in sccm, locate a step called "MDT Toolkit", make a note of the name of the package being used in this step. Close the task sequence and browse to " Packages" and try to locate this package. Once found, right click it and obtain the "source" location. Go to this source location and thats where the MDT files being used in your sequence are. Those are the BMP you want to update, not the ones in your MDT share. Remember to update the DPs once you replaced the images

Most definitely can be done, and the more that you want automated, the more tedious the initial configuration will be (dynamic query collections, uninstall packages, config baselines etc). Youd be pulling together multiple features but could in theory automate the whole rogue software uninstall process. Much depends on what you classify as "rogue" software. If you are not fully co.fortable woth sccm i'd start basic. Go to assets and compliance > inventoried software. Search for names of rogue software like 'torrents', 'java' etc. Will give you a overall view of whats in use in your environment. Once you have names of software you want to uninstall, design a query that lists these names, version and uninstall string of your rogue software. Using the results of the query, you can create a script to uninstall the rogue software in one hit or segregate it. Theres a lot more to this, if you need further info i can follow up. At a high level what you want to do is: - Determine what rogue software is in use - Define a baseline of software you want (latest Java, latest Flash etc) - Deteine if you want instant remediation of noncompliant machines or if you are happy to accept a 1 week or so timeframe. (Hw/sw inventory runs once per week by default whereas a compliance baseline could be set to run much more frequently thus giving you a faster result) - Establish a uninstall mechanism (script, in hours, out of hours etc) - Maintainibility, what will you do when a new Java comes out. How easy will it be to update the exisiting setup to cater for this. Lots of consideration as you can see.

Download treesize or windirstat and run it on your C:\ Get a visual picture of whats taking up the space, may not be anything to do with sccm at all.

Ah, so theres something to work with. Google: "boot image" 2152205056. Lots of varying issues. I'd start with the simplest - Open SCCM console using "run as administrator" - Clear the Windows\Temp folder as sccm uses that to build your temporary WIM before copying it to your chosen location - Try importing a few drivers only - Double check the driver source is all intact and hasnt been deleted or modified If you are not too concerned, there is an alternative way of injecting drivers into your boot image, although personally i would advise doing it via sccm so you can remove driver easily later on. The alternative method is using DISMGUI. It allows you to modify the .wim file directly, bypassing sccm completely. See this for how to: http://www.deploymentninja.com/home/easily-add-drivers-and-updates-to-images-with-gui-dismI used this method at my current organisation as we were gtting various issues injecting drivers and in the end i got fed up and injected them offline using DISMGUI. Works without issues, the only downside is removing a driver (if you ever needto) will be tricky

Firstly temporarily disable AV software on your sccm server. I have had countless problems with McAfee blocking me from creating a Boot Image in SCCM as it detects it as a threat. Everytime I change a boot image I have to disable McAfee (I have been to lazy to put exceptions in McAfee) Secondly, what specific errors are you getting. Error codes are vital for narrowing down issues in sccm. In order to make PXE boot directly to your new boot image, you need to make sure the option of "Deploy this image from a PXE enabled point" is enabled (re-distribute the boot image to the DP after doing this). Then also ensure you deploy a task sequence that uses your new boot image. SCCM uses a "last deployed" rule, so whatever boot image you last deployed, will be the one that gets used by PXE.

Looks like PAT doesn't have enough access to allow you to make changes to accounts (especially make them sysadmin) A lot of people get caught out here, often people forget the password to their management studio and resort to rebuilding. Before doing a rebuild, try this: 1 - Downloaded http://technet.microsoft.com/en-us/sysinternals/bb897553.aspx and extract psexec.exe 2 - Copy psexec.exe to c:\ 3 - Open a administrative CMD window and run: c:\psexec -i -s "C:\Program Files (x86)\Microsoft SQL Server\.................Ssms.exe" Obviously replace the location in the above command, you need to locate where the exe file for Management Studio is located. It should be called ssms.exe and somewhere in Program Files (x64/x86). This will launch Management Studio using "NT AUTHORITY\SYSTEM" and you can then add a account (or modify PAT) to make it sysadmin.

not sure why a client re-install wouldnt work but rather than having domain join at the end of a task sequence, add it BEFORE the sccm agent is installed as normally this is when it occurs. I cant remember off the top of my head exactly where it falls but the step is called "Apply Network Setting" and comes after Apply Operating System but BEFORE "setup windows and configmgr" Do a google image search for: "sccm task sequence network settings" to determine where you should place your domain join step. Domain join is a very routine step that rarely causes issues as you have described. Maybe adding it to the end is causing an issue somewhere.

Sorry to hear your customer is being difficult. Unfortunately this is a tough one to answer and a very rare scenario. First thing i would do is setup a lab environment and try to replicate what you suspected. Set up your lab, trash the "nt system" account and try to get a feel of how sccm behaves. I know you wont be able to replicate the issue like for like but at least youll be able to determine (and prove) that to tinker with "nt system" you need admin privileges which you say you didnt have at your customers site. If possible the logs from the customers site would help too although depending on how long ago this issue occured they may have been overwritten

Sccm 2012 will work fine. Totally youd cboice as to whether you want 2012 r2 or sccm 1511. Depends whether the features it offers and licensing suits your needs. Im pushing to install 1511 at my workplace. Although sometimes the "latest and greatest" can also come with bugs. Sccm 2012 r2 has matured very well and is very stable. Will 1511 be just as stable?. My opinion is yes as its all built on the same technology unlike 2007 to 2012 which was a drastic changw

I cant be sure but you can probably harness the "System Console Usage" or similar hardware inventory information to get you what you want. The downside being that its "stored" information from a database rather than "live" real time info. Occasionally i need to deploy software silently and due to the complexity of the applications i prefer that nobody is logged on (unattended). To do this i add a few lines of code to my app/package which scans if a user is logged on. If nobody is logged on the install proceeds. If a user is logged on the the install fails with whatever exit code i set. Not the most efficient way of doing things i know. Have you considered the built in option to run a install "when no user is logged on"

Interesting topic. We are considering the possibility of a side-by-side migration soon. Figured it gives us more control over migration of clients as we can keep the old site running. My plan was to build a new server, new site code etc and then use powershell to export as much as I can (collections, applications, task seq's etc) and then re-import them into the new site. Although this adds some extra work, we are left with a "clean" site and can tidy up some of the "bad practice" that was carried out by the previous sccm admin when installing.

An alternative to my suggestion above is, create a config baseline that checks if your file exists. Then create a collection based on all non-compliant results of this config baseline. Finally deploy a package to this collection. Voila!! Amazing what sccm can do, just requires a bit of thought

There is a simple solution to this. Create a APPLICATION that uses a "script" deployment type. Set the "detection method" which checks if your file exists. Set the deployment command of the apllication to a simple script that copies your file to the destination you want. Deploy it as Required. So now, sccm will use the detecfion method to check if youre file exists. If it does, nothing happens. If if doesnt, sccm deploys your application (which is just a script that copies your file into your location).