Zenden

That looks quite interesting but unfortunately wouldnt be suitable for our environment. It could be useful for stations for print job releasing etc though so i will keep it in mind.

Bad Address basically means the following precess has occurred: DHCP Client requests IP from DHCP Server DHCP Server responds with an IP for the client Client does an ARP lookup on the address and if it gets a response then it knows that address is in use. It then responds to the DHCP server to inform it that the IP is in use so it therefore wont try to hand it out again. DHCP Client requests another address.

Good Morning All, I spent a portion of the half term break trying to get as much speed as possible into the student logon process in our ICT Suites and found a few useful bits and bobs i thought it was worth sharing (as well as for me to check back on when i inevitably forget what i have done!). I know there are a few (very useful) posts on here regarding this already but i found some things which have not yet been noted so I hope someone finds them useful. After creating and copying the profile to our network share i did the following: Remove all policy entries under HKCU\Software\Microsoft\Windows\CurrentVersion\Policies and HKCU\Software\Policies Remove all startup entries from HKCU\Software\Microsoft\Windows\CurrentVersion\Run and RunOnce Delete all log, blf and regtrans-ms files which have been created in the root of the profile (Adobe CS3) Wipe %appdata%\Adobe\Acrobat\Distiller 8\Cache Folder (Adobe CS3) Delete %appdata%\Adobe\Fireworks CS3\Styles\Style Defaults.stl Delete %appdata%\Microsoft\Windows\IECompatCache\index.dat and %appdata%\Microsoft\Windows\IECompatCache\Low\index.dat Delete %appdata%\Microsoft\Windows\IETldCache\index.dat and %appdata%\Microsoft\Windows\IETldCache\Low\index.dat These tweaks brought the size of our profile down from 7+MB to 1.3MB I also made the following tweaks to the library-ms files in %appdata%\Microsoft\Windows\Libraries by dragging them into notepad and editing their contents: Remove the content of the tag in the library files. If this tag has content then it seems to stop the redirected/user specific folders from being included in the libraries and instead shows the local profile of the user who the profile was based on (eg C:\Users\profileuser\Documents etc) Remove the last tag (both the tag and its contents) in the library files. This is the link to the public documents/pictures/music/videos folders. Removing them stops them from appearing under libraries when users login and just leaves their homefolders. Once this was complete the login times reduced from 90 seconds to around 45-60. (7 year old PCs and 100Mb Switches). I then setup an XCOPY Script to copy the mandatory profiles to the local c drive of the machines as the copying of the profile via the network was responsible for around 21 seconds of the login time. Script pasted below: IF NOT EXIST C:\Installed MKDIR C:\Installed IF NOT EXIST C:\Installed\__LOGS MKDIR C:\Installed\__LOGS IF NOT EXIST c:\installed\Profiles\installed.txt goto install IF EXIST c:\installed\Profiles\installed.txt goto installed :installed ::echo "already installed" goto end :install eventcreate /l system /so ProfileScript /t information /id 1 /d "The Mandatory Profile Batch Script has detected that it has not previously ran on this computer and has begun" MKDIR "c:\Profiles" XCOPY "\\domain.local\netlogon\profiles\DEPLOY\*" "c:\Profiles\*" /O /S /E /H /Y /C >C:\Installed\__LOGS\ProfileScriptlog.txt XCOPY "\\domain.local\netlogon\Deployments\Installed\*" "c:\installed\Profiles\*" /S /E /H :end Quick outline of the script: It checks if the file c:\installed\Profiles\installed.txt exists. if it does then it knows this script has already been ran and finishes without copying anything. If it doesn't exist then it runs and creates an event log in the system log. It creates the folders c:\installed\Profiles and C:\Profiles . It then copies all the mandatory profiles within the DEPLOY folder into c:\Profiles and logs the process to c:\Installed\__LOGS\ProfileScriptLog.txt. Once completed it copies Installed.txt from a network share into c:\installed\profiles\installed.txt to stop the copy job running at every bootup. I then used a GPO to assign the profile c:\profiles\student to the ICT suites. I also did a number of Registry tweaks via Group Policy Preferences to the workstations to increase speed: Created/Changed HKLM\Software\Microsoft\Windows\CurrentVersion\Policies\System\DelayedDesktopSwitchTimeout (REG_DWORD) to 5 to vastly increase the final "Preparing your desktop" stage of login. Changed the IsInstalled REG_DWORD values to 0 and deleted the StubPath values for a number of items under HKLM\Software\Microsoft\Active Setup\Installed Components : {7790769C-0471-11d2-AF11-00C04FA35D02} - Address Book {44BBA840-CC51-11CF-AAFA-00AA00B6015C} - WinMail {89820200-ECBD-11cf-8B85-00AA005B4383} - Web Platform Customisations {89820200-ECBD-11cf-8B85-00AA005B4340} - Windows Desktop Update Login times now down to 27 seconds on a very old Athlon X2 machine and 17 seconds on a new i3. This has been mostly successful and is fully functional for users using the mandatory profile. However, if a user logs in who doesn't use any kind of roaming profile and a fresh local profile is created then the library files are never created for them so windows explorer shows nothing under libraries and also nothing under favorites (where desktop/downloads should be). After a lot of research i have found that restoring the registry items for Windows Desktop Update fixed the Favorites issue and restored Desktop/Downloads however i have not found a fix for the libraries, in the meantime all users have the mandatory profile applied to them when they login in the ICT Suites. So in conclusion, massive increase in login speed but an issue with libraries. Has anyone else been down this route/ have anything to add/ can advise on the creation of library files at login?

This is 100% correct. I had to deal with a network a couple of years ago that had this problem for 9 months (was my predecessor not me) and it is essential to understand you need to tackle the cause of the problem and not be distracted by the machines it is copying itself to. Where is it coming from? don't bother trying to scan and remove it on the machine flagging it up as it will come straight back. Some machine has it actively running and it is spreading via the admin$ share and using domain admin permissions to authenticate. Find the culprit machine, unplug it from the network. Change all domain admin passwords. Remove the infection on any/all machines but especially the culprit. Patch the culprit with kb958644 (I know that by heart after having to deal with this). Finally bring everything back online and use wsus to deadline that patch for any and all operating systems.

Ditto!

Good Morning All, Sorry for the lack of an update, I had a weeks holiday from work (moving house so not really a holiday but hey!). It seems (hopefully) that this issue is now resolved and that it was in fact the "large send offload" setting on the VMs network card which caused the issue. After returning on Friday I haven't seen the issue once and my technicians are reporting that it hadn't happened while I was away either!. So far anyone on HyperV with a virtual DC. Disable Large Send offload for IP4 and 6 on your VMs network card or you will end up chasing your tail like I did! Thanks for the help everyone

Gigabit Card and picking up an address. It can ping the server, nslookup is fine etc. it is the initial bind of the Pc via ldap which is failing.

Hi Richard, The same Pc actually works fine after a reboot so it is not an issue with domain membership or with the group policy setup or OU location as these all work correctly after a reboot (and then fine for a couple of weeks). It is an issue with the PC (which is never the same PC each time) binding the domain during bootup in order to then apply the policies. It is very strange issue indeed.

I did run a gpresult when i was troubleshooting this last week and i believe it came back with no RSOP data. But i will double check as its been a hectic week and my head is a bit all over the place. I will keep this til last in my troubleshooting as a final "trial and error" step if i cant narrow it down any further, good idea Ive done the diag tests although I havent run a replication test so I will get doing that Time on the client seems fine but they are quite outdated machines, perhaps the time in the cmos could be related? Unfortunately BGinfo fails (i actually use it anyway) as the script runs through group policy! Perhaps running it through the AD profile tab will force the issue though! Much appreciated advice all :-)

Yes, I have one physical DC and one Virtual DC on a hyper-v cluster. Annoyingly enough the error log does not flag up which dc it tried to bind with (this would have helped a lot!). They both have DNS running. At a guess I would say the virtual DC is the problem due to it sharing the physical NIC on the HyperV host (hence I disabled Large Send offload which is described in a similar problem at Intermittent failures apparently related to DNS ). Can anyone recomend a good troubleshooting method to narrow this down?

Good Afternoon All, I am troubleshooting (and raging at) an issue with group policies not being applied intermittently in our ICT Suites. Symptoms: PC boots up and fails to perform an LDAP bind Then applies no group policies to the PC then applies no group policies to any user who logs in! Error Log: System - Event ID 1006 - "The processing of Group Policy failed. Windows could not authenticate to the Active Directory service on a domain controller. (LDAP Bind function call failed). Look in the details tab for error code and description. ErrorCode - 85 ErrorDescription - Timeout Notes: Restarting the PC completely resolves the issue as the bind then works at the next bootup and group policies are applied The issue happens to one computer in 30ish so is proving very dificult to chase! The issue does not happen to the same PC each time Things I have already done: Checked, double checked and treble checked our DNS infrastructure, all is perfect. Disabled Large Send Offload on the Virtual NIC for our virtual Domain controller (one virtual and one physical) I am completely baffled, can anyone shed any light on this? any help hugely appreciated

you may be stuck with emerge then i am afraid. Having said that, I left the school with bromcomm about a year ago so things may have changed/advanced. The user on here who is at that school now is AChandler, he may be able to give you more info!

its just a website so yeah as long as you publish it externally and setup your firewall rules etc.

Bromcomm webfolder has an addon product for ipod touches. my old school had every teacher doing registers on their ipod touch and it worked great. Uses sims account to login too!

ive done the side by side migration recently and it was all reletively smooth. There is some great guides for this on msexchange.org which cover the main pitfalls. Well worth a read!

WSS/Sharepoint and the salamander Web Parts would also be an option, I use the web parts here and they work perfectly. second the point on having the wifi on a seperate SSID and VLAN, could be nightmare if not!

Judging by the fact that is shown in the gpresult, the GPO is correctly linked however it must be configured slightly wrong. Please check the event viewer for any event logs. My guess would be your wording/syntax in the GPO is slightly incorrect causing the registry permission to not be applied. My first guess would be the syntax of the actual registry key itself. keep tweaking it until the permission gets applied. You could try installing the software on your workstation, then using the RSAT to create the group policy on your station aswell thereby making it easier to compare things and point to the correct location in the registry. I have used this kind of setup to apply permissions to upgrade sims and it works perfectly so you will definately get there is you persevere!

4Gon or WifiGear were the only places we found them.

I have dealt with this on a wide scale. Make sure all Pcs and servers have the patch KB948644 (havent double checked that but pretty sure it is right, i had to install it enough times!). If all of them do then it is an inactive infection being pulled either from autorun or task scheduler and wont be causing any damage. Make sure to disable autorun through group policy. Biggest thing to know is that if the PCs are patched then the infection is null and will not spread further, however if any PCs dont have the patch then it will spread using the admin$ share by keylogging passwords of anyone who logs in.

I agree with AngryTechnician, GPO is the best way to do it. Can you confirm you have the syntax and location of the Registry Key absolutely correct in the GPO? is the GPO showing in GPResult? if so, is there an error log stating why it hasnt applied?

Hi Clareq, can you confirm the exact name of the gpo you set this in and its location? i have imported the chrome GPOs but i cant seem to see a setting for this (perhaps my eyes are playing tricks on me...) Thanks in advance.

Thanks for all the info guys. I have been sent a quote for the Starwind CDP version (as i will be using one Storage server rather than having the Storage highly available) and it is coming out at a relatively reasonable 600ish quid. This is to use it with my five node Failover Cluster. This would of course leave the storage/Starwind server as the one point of failure but i do have a strong backup schedule and plan in place (and our current setup has the san as one point of failure anyway). I have tried the microsoft iscsi target and it just doesnt seem sturdy enough for me to trust it but the starwind software looks very good indeed. Looking at around 3000 for a storage server (6TB of storage space) and 600 for the software. comes out at half the cost of your average off the shelf san! can anyone see any problem with this that i havent thought of?

Is anyone using Starwind as a SAN for a HyperV cluster? If so, how is it holding up for you? I currently have a cluster setup with three hosts but the SAN is pretty aged now and I am fairly certain it will die within the next year. I dont have enough in my budget to purchase a proper SAN (I have begged the Head for some more and am also looking into Leasing Options) so I am looking at a plausable PLan B. IBM Server - x3650 M4 with 6x 1TB 6GBPS SAS Drives (for storage), 2x 148GB 6GBPS SAS Drives (For OS) and a redundant PSU. I am hoping to purchase this, add starwind into the equation and have it function as cluster storage for my hyperV hosts. The idea is to get it running for the next year or two, by which point i may be able to afford a proper san and could then turn this IBM into another Host for the cluster Anyone got any comments/experience?

ICT Technician wanted at my old school. Network Manager is a good friend of mine and very highly skilled. Details available at: Vacancies