The Windows Update 0×80070005 error is common on Vista machines and somewhat easily fixed (just Google for the error; plenty of results). Rarely have I seen it mentioned in forums and blog posts relating to a Windows Server 2008 system. We’ve manually installed critical updates on the affected system every few months to keep it current - which is a pain. Service Pack 2 failed to install as well, leading us around in circles with System Update Readiness Tool requests and so on.

We’ve just installed a new Dell PowerEdge R610 - Windows Server 2008 Standard x64 SP1 - to act as a primary domain controller, which was performing Windows Updates just fine until we added the Active Directory Domain Services role and ran DCPROMO. After we rebooted, Windows Updates stopped working and presented with the dreaded 0×80070005 error.

The last update to be applied was KB955430, which:

Updates to the Windows Vista and Windows Server 2008 installation software are included in this update. The installation software is the component that handles the installation and the removal of software updates, language packs, optional Windows features, and service packs. This update is necessary to successfully install and to remove Windows Vista SP2 and Windows Server 2008 SP2 on all versions of Windows Vista and Windows Server 2008. This update is not necessary to successfully install the service pack if you install the full file version of the service pack. The full file version of the service pack includes this update.

I have a hunch that a combination of Win 2008 SP1, ADDS, the KB955430 patch, and the system being a domain controller (so; after installing ADDS, running DCPROMO and rebooting) cause the Windows Update mechanism to fail.

As this new R610 is not yet in production, I demoted it using DCPROMO, rebooted, but kept the ADDS role installed. The server now queries Windows Update and our internal WSUS server correctly, AND lets us install SP2

The 0×80070005 error is resolved! Updates are located correctly via Winows Updates and WSUS and install correctly. SP2 also installs correctly; whereas before it would not.

Unfortunately, as soon as you re-promote the system to a domain controller, Windows Updates fail again with the same 0×80070005 error.

Frustrating!!

Interestingly, Wikipedia’s page on Danger shows their ‘Disaster’ header, which I think is a bit over the top. A ‘Stupid Sysadmin’ header would be more appropriate.

BGInfo (populates the desktop background with system information) can assist with helping to identify the server, however if you’ve seen one BGInfo you’ve seen them all.

I came across Terminals this morning while I was looking for plugins for Microsoft Dynamics CRM. It lets you manage multiple RDP sessions, keep a list of hosts you regularly connect to (no more relying on most-recently-used lists or icons on the desktop), and comes with a swag of handy network tools all built in.

In a workplace that is all about becoming more efficient, Terminals is a huge improvement over the regular mstsc.exe experience.

Hit up here: http://www.codeplex.com/Terminals

It’ll let you track your vehicles by polling the GPS grid every few seconds, then uploading the data to their servers via the mobile phone network. If you’ve got no mobile coverage, the unit queues the data until there is sufficient signal to transmit, and then uploads. It’ll cost you an initial $450 for a single unit, but that provides a years worth of tracking, and then it changes to a yearly subscription service of $90 every year thereafter.

There are comprehensive reports available too, such as travel outside of specific time frames, travel outside a specific radius, and travel over a certain speed, together with the usual plotting of routes, directions, altitudes, and speeds.

Check it out!

As I’m learning up on a whole suite of new technologies at the moment, like VMware’s offerings, and Symantec’s security and backup technologies, I find that most of the information I already know due to experience. Of all the free online presentation/quiz style “entry-level” certifications, I enjoyed VMware’s the most, but the next step up is costly.

Being an IT professional, I don’t have five days to set aside my regular duties and engage in a classroom style course, and being a small business owner in a slow economy, I don’t have the massive amounts of cash required to actually sit the courses anyway.

Even if the funding wasn’t an issue, justifying the expense is difficult, considering that it won’t be long until that course and information set has been superseded.

So, I have a few questions for the community:

• What are your experiences with vendor certification?
• Do you think the costs are justified in most cases?
• Which vendor gives the biggest bang-for-your-buck in their certification (think not only up front cost, but longevity of the product, possible returns for the business, etc)?
• What do you think of hardware certification?

I have just cruised through examinations and earned myself a VMware Sales Professional accreditation, as well as the company now holds the title of VMware Services Provider Partner. Yay us!

This is the first of many new certifications and accreditations to come, as well as many new partner programs.

Look out Dell and HP!

• install SBS2008 on a new server
• convert the SBS2003 to a virtual machine, wipe the existing server, install SBS2008 on existing server.

We went down the road of the second option, and overall the install process was satisfactory. We did run into some snags, as during the first attempt of installing SBS2008 the process crashed. The following attempt the server seemed to get ’stuck’, waiting for response from the SBS2003 VM.

The problem in the second attempt was that the 2008 install had transferred the domain operations master, infrastructure master, schema master, and other roles away from 2003 in the first install attempt. When the second install tried to do this, it could not complete this step and was stuck. The 2008 installer did not give any reason for the problem, so we had to manually figure it out.

The roles were seized back to the 2003 server using NTDSUTIL, and the next install attempt was much more successful!

I am very impressed with the SBS2008 system. We manage our client’s SBS boxes, rather than letting them do it, so even from an experienced point of view, SBS2008 is really quite good. I don’t like how WSUS is automatically enabled; I think this should be one of the ‘Getting Started Tasks’. I do like how easy it is to relocate an Exchange store, WSUS data, shared folders, etc, to another drive via the SBS Console.

Once running, configured, and stable, VMWare Server v2 was installed, and a Ubuntu linux virtual machine installed as the first VM on the system. With the Ubuntu VM we run an in house lab for PHP web development. I’m undecided about Ubuntu, after spending five years with FreeBSD. Our first project on this lab VM is a redevelopment of our client management system, something which has been on hold for about a year. More on the project to come in later posts.

In the past six months I have gone through two serious (one of them majorly serious) failures relating to the same server. I’ve noticed there are two good things to come out of a serious server crash/outage. When I say serious, it is on the level where it impacts the majority of the organization for longer than 24 hours, however only time is lost, all company data is safe.

The first positive outcome is that those who are responsible for maintaining and bringing systems back online get invaluable experience. You can’t train for it or study for it, because in a lab environment you don’t have the bosses and staff breathing down your neck. For systems where there is little or no redundancy (probably due to budgetary reasons or the management-imposed requirement to K I S S), the pressure starts to build.

Trying to get the system running again, diagnosing the faults for vendor tech support staff, keeping management up to date, while answering the calls and the same repetitive questions from the userbase can make for a stressful few days. Especially when the exact same server crashes 80 days after it was rebuilt from the first crash.

As the pressure and wariness from long days build, the likelihood of rushing through processes, thereby making mistakes, increases, in turn increasing the time it takes to get the system back online.

The experience gained in these scenarios is invaluable; you can train and study for recoveries, but you can’t get the experience without being put through the wringer.

The second positive outcome is that the majority of business managers will reassess their IT systems and be more willing to act on suggested improvements or increase the budget. Because the managers become more aware of the impact of downtime on the business, they will likely consider the recent proposal from IT to implement clusters/load balancing/replace hardware/etc as worthwhile, rather than the IT team asking for more toys.

These recent crashes have cost the client around $8000 in IT labour, which is not huge, but it has cost them many times more in lost productivity. In life, hindsight is 20/20, but in business, foresight is generally listening to the IT department’s recommendations..

For the new Exchange box, we ended up with a Dell PowerEdge 2950 III with two RAID 1 arrays, dual quad-core, 4GB RAM, etc. It arrived about a week earlier than scheduled, and so far has been a breeze to work with.

Installation into our rack was a little troublesome due to the inability of the server rails to extend the 2-3mm extra we needed. We have a 1000mm deep rack from Cheval with the front posts moved back 70mm, and the rear posts as shipped from factory. The Dell rails were slightly too short to fit perfectly, however an adequate fit was achieved by some extra turns of the screwdriver.

It’s been about a week since the server was installed, however only just yesterday Exchange 07 was installed on the box. Installation was without trouble, and much more informative than the Exchange 2003 installation. My previous experience with Exchange 2007 is restricted to labs and virtual machine’s, so this is the first planned-for-production server that we’ll have implemented.

As of the date of this post, Symantec’s Mail Security for Microsoft Exchange does not support Exchange 2007 on a Windows 2008 x64 Service Pack 1 system. After installing SMSMSE and finding that it had DESTROYED the Exchange installation (Microsoft Exchange Active Directory Toplogy Service would not start, subsequently none of the other Exchange services would either), we had no option but to format the box and start again.

I am very impressed with the ease of use of the Exchange Management Console. It seems to make more sense to me than the Exchange 03 management tools, but as some of the major items have been re-worded, it can be a little overwhelming when you’re under the pump to put it into production.

Mailbox migration was very easy, albeit slow because of the faulty SBS2003 box hanging every few minutes (4 hours to migate a 4GB mailbox??), however once done there was very little we had to do in the way of getting clients connected up. Infact, Outlook 2007 took care of this process entirely but prompted us with SSL certificate warnings.

The SSL warnings were easily fixed - Microsoft KB940726 fixed the warning when opening Outlook, and although we had it about 85% correct, we finished off with a quick review of Sembee’s article titled Exchange 2007 Single Name SSL Certificate, and the remaining glitches were corrected.

We’ve been in production for just under two weeks, and the only issue reported was a SSL cert warning from one of the staff’s home computer - saying that “The security certificate presented by this website was not issued by a trusted certificate authority.” A root certificate update on the system should remedy the warning - this has been advised but the outcome is yet unknown.

The previous SBS2003 box was cleaned up a few days later and a vanilla Windows Server 2003 was installed. So far the system is running as expected - fast and trouble free. It now has the O/S on a RAID 1+hotspare.