Another quick post to show how you can quickly deploy fully working DHCP server with multiple scopes in a matter of seconds. In my case it was a single server with 90 very different scopes and doing this manually would be just soo boring and long that’s unreal. Unfortunately Windows Server 2012 wasn’t an option so no PowerShell love but Windows Server 2008 R2 is still pretty decent and using netsh wasn’t as painful as it seemed. To get us started we need to install the DHCP Server role and start the required service (dhcpserver):
sc config dhcpserver start=auto
net start dhcpserver</strong>
Next step is to authorize DHCP server in the enterprise so we can actually use it to dish out IP addresses:
I use Windows Server 2008 R2 SP1 DNS to resolve my public names and precisely every two days I get “Can’t find google.co.uk: Server failed” messages and the usual “Internet Explorer cannot display the webpage” nonsense. This only happens when using root hints and not forwarders and only for few selected top level domains (.co.uk being one of them) Here is what the situation looks like when the DNS server is broken:
Ping returns “Ping request could not find host google.co.uk. Please check the name and try again” and nslookup throws error in title. Pretty annoying and so far the fix was to simply bounce the dns service – not elegant in any shape or form but highly successful! Continue reading →
One very useful piece of information to know, if you’re working in large Active Directory implementation with multiple DC’s and Sites, is to be able to determine which Domain Controller machines are authenticating against at any given time.
Here’s the command to tell us exactly that:
Rather handy if you’re testing whether Sites and Services have been setup correctly or are moving computer objects in Active Directory and want the quickest turnaround for GPO’s etc.
You could always use built-in ‘set l’ command but that’s not always accurate due to %logonserver% variable taking a bit of time to update if you change Active Directory site and ultimately domain controller you’re authentication against. Remote workers are good example as they tend to be all over the place and ‘set l’ might give misleading and not precise results.
Deploying software in the enterprise can be quite complicated, it doesn’t have to but it can be at times. There are tons of people that will just “install” what’s necessary to get them out of trouble (even if it means doing it 100 times over and over!) The same group of people cannot really distinguish between installing and deploying and if you happen to have them on board – you’re in trouble. Now, I don’t even remotely consider myself software packager but when deploying applications I usually follow few simple rules to take care of the basics. Here are the rules:
I. Installation has to be silent with no user intervention required. You run the install and usually don’t even know when its done unless you watch out for msiexec processes etc.
II. Once installed you should run it and no popups / random windows / activation, registration dialog boxes etc. should come up. It simply has to open nice and clean.
III. Again, once installed no extra rubbish should be copied to users desktop / start menu / run once etc. No auto update services should be allowed to start. No tray icons should be visible etc.
If you’re able to achieve all of the above you’re better than most. What we are essentially looking for here to clean and sleek deployment of whatever we need to push out to our client machines. No one likes to login to a machine and see 10 popup windows asking you do to do stuff. Horrible experience before you even start doing anything. Lets get to it!