When trying to access Windows Components in Add Remove Programs, you receive the following error: Setup was unable to open information file fxscom.inf.

The problem is that you are using an XP build using nLite. NLite removes uncessary Windows components to make a compact and speedy image. Unfortunately, it is not officially supported by Microsoft. When you install Windows XP Service Pack 3, the sysoc.inf file is overwritten. The deleted configuration information for the components is put back into the new sysoc.inf file written by Service Pack 3. When you attempt to open Windows Compoents in Add Remove Programs, it will look for the missing pieces due to sysoc.inf.

To fix this, I did the following: as each error comes up in Add/Remove programs, simply delete the corresponding line from sysoc.inf. Make a backup of sysoc.inf first. Sysoc.inf is located in your windows\inf folder (maybe hidden, so enabling viewing hidden files or be a real man/woman and use the command line <giggle>.

The error will also result in errors for the following files (components):

xsocm.inf
fxsocm.dll
setupqry.inf
setupqry.dll
fp40ext.inf
fp40ext.dll
msgrocm.dll
msnmsn.inf
rootau.inf
games.inf
communic.inf
optional.inf
pinball.inf
igames
zoneoc.dll

As long as I’ve been in IT, users have demanded the capability to email files. It is tremendously convenient but leaves much to be desired for IT personnel in charge of messaging systems.

Typically email attachment sizes are limited. Users complain when attempting to email huge spread sheets or media content if they are denied due to size restrictions.

Typically the solution has been FTP. However FTP generally leaves much to be desired for the average user. It requires the maintenance of accounts. Some people are restricted by local policy and firewalls. Browser support for FTP is weak. FTP servers must generally be maintained or entropy sets in and they become an administrative and security nightmare.

Enter Allard Software’s Filetransfer Appliance.

It works like this: When a user has the need to send a large attachment they point their web browser to the File Transfer server. They get a page with a form to send an email. They fill in To:, Subject:, and their message. They then browse the file and click upload. Then they can send the email. The recipient gets the email which contains a link to download the file. The file expires after a set period of time. The recipient receives a notification when the file is downloaded. That’s it! Beautifully efficient and simple!

Filetransfer Appliance is a piece of cake to set up. It installs as a VMWare appliance - meaning the software is already installed and configured. All you need to do is boot the VM, give it a hostname and set your IP address. There are a few configuration options including your mail server and Active Directory integration so users can authenticate with their AD credentials.

The device can use SSL with your own certificate or a self signed certificate if you choose.

Support is highly responsive from the developer himself.

The device is incredibly inexpensive, licensing starts under $100 for 10 users and their is a FREE 3 user demo.

http://www.allardsoft.com/filetransfer

I stumbled a great post over on AtomicVinadloo about using Tor and Privoxy on Ubuntu. (or any Linux distro). I had to take an extra step to actually make it work through Privoxy. (I think his config has it connecting directly through Tor and bypassing Privoxy).

Basically Privoxy is a proxy server - you can configure it for all kinds of filtering options to customize your browsing experience, block ads, etc… Tor anonimizes your connection to the Internet. It works by making an encrypted connection to a server in the Tor Onion network. Once in the network, you are routed through random Tor servers. The result is that the end web site cannot identify your location or details. Your traffic also remains encrypted between your computer and the Tor network - meaning nobody at your ISP is able to see your traffic.

When Privoxy and Tor are used together they provide a great deal of control. Tor can be used without Privoxy as well. To be fair, there is also a version for Windows.

Configuring Privoxy and Tor in Ubuntu is a piece of cake.

1. install Privoxy and Tor:

sudo apt-get install tor

sudo apt-get install privoxy

2. install FoxyProxy plugin for Firefox (this is not absolutely necessary but makes it easier to switch Tor on and off.

https://addons.mozilla.org/en-US/firefox/addon/2464

Do not do any auto configuration - just cancel through any prompts to configure for now.

3. Edit your Privoxy config file

sudo gedit /etc/privoxy/config

uncomment (delete the #) the line that says:

forward-socks4a / 127.0.0.1:9050 .

and save

4. Restart Firefox and right click on FoxyProxy in the status bar

Click “Add New Proxy”

Enter the following:

Under General;
Enabled: Tick
Proxy Name: Privoxy To Tor
Animate ….: Tick
Include …..: Tick

Under Proxy Details;
Manual Proxy …. : Tick
Host Name: 127.0.0.1 Port: 8118
SOCKS proxy? : do NOT tick

Under URL Patterns;
(If it doesn’t exist already)
Add new pattern;
Enabled: Tick
Pattern name: All
URL pattern: *
Whitelist: Tick
Wildcards: Tick

5. Restart the Tor and Privoxy Services

sudo /etc/init.d/tor restart
sudo /etc/init.d/privoxy restart

6. Go to http://whatismyip.com and note your ip address.

7. Right click on Foxy Proxy and enable your new “Privoxy to Tor”

8. Refresh the whatismyip.com page and note your new address. This confirms you are now browsing via Tor.

You can also confirm you are browsing via Tor by going to the following URL: http://check.torproject.org

For more information on using Tor with Windows and different ways to use Tor and Privoxy, visit the following:

http://www.torproject.org/

http://www.privoxy.org/

You will find that the speed of browsing leaves a bit to be desired - however in the case of NGO personnel in hostile countries, or other sensitive matters, it can be an invaluable tool to protect privacy.

We just upgraded our ESX farm to VC 2.5 ESX 3.5. It was on an older 2.x version of Virtual Center. The ESX hosts were 3.0.2.

We then realized our SSL Certificate (The VMWare provided, self signed one) had expired a few days prior. Our upgrade to Virtual Center 2.5 and ESX 3.5 did not replace the expired certificate.

We looked into the possibility of installing our Go Daddy wildcard SSL certificate, which requires a process of conversion and it didn’t seem worth the hassle to us. See the links below for more on that.

We were able to rename c:\Documents and Settings\All Users\Application Data\VMware\VMware VirtualCenter\SSL\rui.crt to rui.crt.old and then go to Add/Remove programs and reinstall Virtual Center Server with the Repair option. This installed the new certificate for us.

At this point, all the hosts were disconnected, but right clicking and choosing ‘reconnect’ brought them all right back into VC.

The certificate is now valid for two years.

Install Wildcard Certificate on Virtual Center
Installing a New Certificate

Why is it that I can Google nearly any event log error or ID in Windows and come up with at least a few references, yet when I use the built in links to research the error with Microsoft at least half the time I come up with, “No results were found for your query. Please see Search Help for suggestions.”?

Some things are just plain stupid.

I’ve been using the free Solar Winds Exchange monitoring tool to keep an eye on my Exchange 2003 backend. Solarwinds has released a free tool to monitor ESX servers. Note this free tool will only monitor one server - if you want the full deal you have to pay for the monitoring suite. These instructions will get you started with the free tool.

1. Download the Solarwinds VMWare ESX Monitor.

2. Install it in your Windows workstation.

3. SSH or console into your ESX server as root, or equivalent.

4. type nano /etc/snmp/snmpd.conf to open the configuration for editing.

5. make sure there is a statement that says rocommunity public, if not add it. Otherwise close.

6. at the command line type: chkconfig snmpd on t (to make the service start automatically)

7. Restart the snmpd service: service snmpd restart

8, open your ESX firewall for snmp port 161 UDP and TCP by entering the following:

esxcfg-firewall -o 161,tcp,in,snmp
esxcfg-firewall -o 161,udp,in,snmp

Now open up your Solar Winds monitor and enter the name of the ESX server. The defaults should work fine unless you changed the public string name to something else.

I’m not a fan of Outlook. I think it is bloated, slow, and prone to any number of issues. That’s me. I like a nice, compact, fast email client that lets me get through all my email quickly. I’ve always liked Outlook Express and wondered why more people don’t pay any attention to it.

I work in linux a lot as well these days. I’m just finding after years of Microsoft that I really like linux. It is my base OS and I run VMWare or RDP sessions to manage my Microsoft shop.

I’ve been wanting to use a simple POP or IMAP client to check my mail when I’m traveling, have poor connectivity, etc. My company has many people overseas on very slow and highly contended VSAT connections. I don’t like passing my credentials in plain text.

I applied a secure certificate to the virtual POP3, IMAP, and SMTP servers on our Exchange 2003 front ends. I opened the appropriate firewall ports, and voila we had secure IMAP and POP3. SMTP was a little weird.

Exchange does not use a separate port for secure SMTP. It uses port 25 for everything. This works out fine. When you configure your client, choose TLS and not SSL. Make sure it is port 25. You can confirm your email was sent encrypted by looking at the header (send yourself a test). It will say the email was received by your smtp server in an encrypted session.

One last gotcha - the secure SMTP worked inside the firewall and not outside. When using telnet into the mail server on port 25 from the inside, an EHLO issued a full string of options, including STARTTLS. Outside the firewall, these options were only a string of XXXX’s. Cisco firewalls using inspect ESMTP statements filter out the STARTTLS option. This also causes the client to fail with an error stating the STARTLS is not offered. Remove the ip inspect esmtp statement and all will be well.

Outlook and Outlook Express use SMTP 25 and the SSL option, not TLS. Any other client, choose TLS.

Clients tested: Thunderbird, Evolution, Outlook Express

Note: if you are using Outlook or Evolution you might consider using the rpc/https built into these two clients.

Recently I tried to export to a .csv from Outlook. Outlook kept crashing. I tried on another machine, same thing. I had another admin try on his computer, and his crashed too. I found a lot online about software conflicts and plug-ins. None of these applied. I tried Outlook safe mode, still crashed.

The answer was to uninstall Microsoft Exchange Tools and run the command fixmapi.

Now, it works.

AppName: outlook.exe      AppVer: 12.0.6316.5000     AppStamp:4833a470
ModName: oladd.fae      ModVer: 12.0.4518.1014      ModStamp:45417457
fDebug: 0       Offset: 00008c24

Windows XP SP 3, Outlook 2007 SP 1

I’ve had my own sites for quite a while and been through many experiences hosting my sites with various hosing companies. By and large you get what you pay for. Many are very cheap, and if you can do it yourself, they provide a very inexpensive way to maintain your little piece of the Internet.

About a year or two ago I started hosting my site on Bluehost. They had a great reputation as being cheap but very reliable. They were considered by many to be the best under ten bucks a month hosting on the Internet. My experience was pretty good for a long time. I never needed to contact support. Everything pretty much just worked out of the box. That is until the past few weeks.

I was used to getting emails from Bluehost from time to time. The owner himself would even send emails after some event occurred that had affected service or after an upgrade. I always thought that was a nice touch. I’m not running a business. Although my email is important to me I can almost always tolerate a few hours of downtime in an emergency. Bluehost’s guarantee is triple nines, or 99.9% up time. That means about ten minutes per week downtime at most.

I received an email that maintenance was to be performed on the server hosting my account. The email stated that the maintenance would be performed on a Sunday night and the outage would be about two hours. No problem - and thanks for the head’s up, right? What followed was that my account was on and offline at random for nearly two weeks. I was pretty patient in the beginning. Things happen. I work in IT. I can give a brother a break. But the outage went on, and on, and on…

The original notification said that they were doing an “upgrade” to consolidate servers and provide better service to their customers. I manage a server farm, including a virtual server infrastructure. The main motivation in ‘consolidating’ servers is ease of management for IT. It reduces operating costs. It isn’t an upgrade - other than perhaps enabling IT to provide better management or resources to the end user.

What I discovered was that the ‘new’ server hosting my account had a CPU that was maxed and it was out of disk space. This was my upgrade. I was only intermittently able to access my email or websites throughout the day. This continued for two weeks.

At no time did I receive notification that there was a problem. Bluehost’s network status paged showed green lights. I was never notified.

After many fruitless emails to support I was finally responded to by a level three engineer who explained the “upgrade” had gone badly and they realized there was a problem with the RAID controller on the new server. He explained that they were over-nighting a new one and it would be fixed the next day.

Ok, great. Here is a professional server farm, hosting hundreds of accounts, and it takes twenty four hours to put a customer back online because of a dead RAID card? What happened to the magic of virtualization? What happened to having homogeneous hardware and a stock of parts? What about reversing the migration and putting my whopping one GB of content back on the original server?

When I asked Bluehost if they planned to make any compensation to their customers they said no. I asked about the 99.9% guarantee. Do you know what they told me? They told me the 99.9% guarantee only applied to their entire infrastructure as a whole. The only 99.9% percent guarantee was that their entire operation would never go offline. It had nothing to do with the service provided to an individual customer.

The response from level 3, although appreciated, was a day late and a dollar short. I had recently paid up with Bluehost for an entire year. I didn’t care. I had to move my account on a matter of principle. I’m trying to get a credit for my remaining months now.

The next day I signed up with LiquidWeb and spent most of the day this past Sunday migrating over.

So long Bluehost. What a drag.

This is one of the funnier videos I’ve seen in a long, long time. I got a very well needed laugh from it. If you’ve ever worked in IT, you’ll enjoy it.