Time and time again this point has been driven home to me, the best way to increase your DR budget is to actually experience a disaster. Now on a very high level everyone knows that a disaster could occur in their environment but it doesn’t usually seem to hit home with some people, much in the same way smokers on some level know they could get lung cancer but they still smoke. As a network admin we can simulate failures and draw up reports and recommendations for management and sometimes that drives the point home and sometimes it doesn’t but nothing ensures you an instant DR budget increase like actually experiencing a full out “losing money by the minute” disaster.
The problem for some companies is that one disaster is all it takes to put them out of business and they never get the chance to fix their ways.
So what can we do as admin’s? Run DR model’s, simulate failures and make sure that your vulnerabilities are well documented and reported up the chain. Hopefully you get the budget you need and if you didn’t well then at least your posterior will be covered when the excrement hits the fan.
I was notified by some end users that they were unable to access shares on a server, they were getting insufficient system resources. I logged onto the server and got the same error trying to access the server’s own shares.
This error was in the event log
“An I/O operation initiated by the Registry failed irrecoverably. The Registry could not read in, or write out, or flush, one of the files that contain the system’s image of the Registry.”
Rebooting the box fixed the issue but here is the weird part. I did a proper reboot of the system but when it came back up it asked me to specify why the server shut down unexpectedly.
After posting 11% drop in profit MS announces layoffs of five thousand staff. Lets put this in perspective though. Drops in profit do not mean losses. MS still posted profit of 4.17 billion dollars for the 3 month period.
4.17 billion profit in three months … are we supposed to feel bad for them now? In the midst of economic uncertain times when many companies are struggling to survive I would think this could be considered pretty good results.
MS has also announced that they will be suspending merit salary increases for the year. The cynic in me has to wonder if this will include the exec’s or just the common workers who I suppose might accept the fact they are not getting a raise with the feeling that they should be happy they even have a job.
I wonder if MS thinks that cutting out merit increases will encourage there workers to work harder during these uncertain times or perhaps they think that fear is motivation enough.
The beatings will continue until morale improves.
Microsoft just released a beta version of virtualization software designed to run legacy (XP and 2000) apps better within Vista. This has been released no doubt as a stop gap measure to appease users with never ending Vista headaches until Windows 7 comes out.
Here are two immediate issues I see
1) Installing a MS app to fix another MS app is like trying to cure your cancer by injecting yourself with HIV
2) The question isn’t whether your applications will suffer a performance hit, the question is how much.
If you need to run XP applications so badly wouldn’t the better option be to not install Vista?
Here is an article about it.
I got this error while troubleshooting an enviroment with two DC’s that were not replicating. This error popped up when I tried to manually replicate in AD sites and services.
I used the repadmin tool to troubleshoot and found that the server had replication turned off.
Enabling replication resolved the issue, here is the syntax
repadmin /options *servername* -disable_inbound_repl
repadmin /options *servername* -disable_outbound_repl
There are times when a DC has to be removed by force. It’s really not that hard a quick dcpromo /forceremoval will do it for you. Now comes the tricky part, you have to remove the metadata manually out of AD.
Here is a really great article for this:
This becomes doubly important if you want to ever re-promote the server back to the domain. Without removing the metadata it will not allow you to re-promote.
Without Virtual Center it is possible to clone a VM though there are a few more steps.
First create the target folder in your datastore, log into the command line and use “SU -” to log in as root.
Use the command “vmkfstools -i \vmfs\volumes\”datastore name”\”source folder name”\”source file”.vmdk \vmfs\volumes\”datastore name”\”target folder name”\”target file”.vmdk
Once the clone is complete you can create your new VM, point the disk source as the new vmdk file and Bob is your uncle!
The ability to ping from a firewall is a fantastic and some might say essential tool for troubleshooting network (in particular routing) issues. It is for this reason that pretty much every major commercial firewall product out there has this ability (even Sonicwalls). Yet for some reason the Watchguard Edge’s do not. Why this is I have no idea, I could speculate but it wouldn’t be very complimentary.
As a consultant I am more often than not working on servers that I myself did not build nor had any input on. One of the things that drives me crazy is when there are services running that are set up under the administrator log in. What this means in practical terms is that before I can change the admin password (which is a good practice to do on a regular basis) I have to log into each server and check what services I may or may no be affecting. I don’t need to tell you that I have much better ways I could be spending my time.
Since installing WSUS on a DC I have been getting intermittant error “Faulting application w3wp.exe, version 6.0.3790.3959, faulting module webengine.dll”. After doing some research I found a pretty easy fix.
Hope this helps