ScriptAlias

OS: Ubuntu Server 8.10
VMWare Server 2.0

Having spent time shoe-horning a couple of virtual servers into an Xenserver environment whose license preceded my current employment, I finally gave up and tried VMWare Server. What a relief! I think we originally spent about $3000 on Xenserver for a couple of perpetual licenses.

VMWare Server was free and worked almost immediately out of the box. It had a few hiccups. The largest was in cloning virtual machines. It’s not supported natively. After a lot of trial and error with manually editing files I finally found a script that did most of the work for me with just one major exception: it didn’t provide my newly created clones with a working network card.

I downloaded the scripts to clone and copy (I rely on the cloning script only) here. I don’t recall what search string finally got me there but I sure was happy to find it.

Here are the files in case the link above fails:

change_name.sh (1.1 K)
clone_vm.sh (1.2 K)

A few things about the VMWare Server install should be noted. I first installed Ubuntu Server 8.10 for my host. Then after I downloaded VMWare Server, I couldn’t get it to install until I issued:

apt-get install linux-headers-`uname -r` build-essential xinetd

Until then, I kept getting compile errors. After I had VMWare Server installed it was pretty straight forward to install virtual machines. But what about cloning?!!

I successfully cloned a virtual machine with from a directory inside the host machine (the host machine is what I installed the VMWare Server in):

sh clone_vm.sh alpha beta

I was then able add a virtual machine to my inventory from the VMWare Server Infrastructure Web Access Page. But even after doing this I couldn’t get my second server, the one that I had cloned from the first, to start a functioning network card. The issue was due to MAC addresses. To fix this issue I did the following:

1. Edit the Network Adapter in the VMware Server console so that it has a manual MAC address

2. Edit the following file in your virtual environment so that the mac addresses your manual MAC address from above
#nano /etc/udev/rules.d/70-persistent-net.rules

3. Edit the following file so that ethernet0.generatedAddress = “manual address from step 1″
#nano /var/lib/vmware/Virtual Machines/new_virtual_machine/clone.vmx

4. Edit your network interface in the virtual environment if you want a static ip (you can use DHCP)

After I restarted the network from within my new cloned virtual environment I had a totally accessible copy of my first hardened virtual machine. Saved me tons of time setting up new virtual environments.

Note: One my next try after posting this, I had to reboot the Virtual Machine to get these setting to stick.

Quick note to self. When installing PHPMotion on newly installed Ubuntu 8 Server there are four very important things to do.

1. PHPMotion has a cgi-bin directory. Set apache2 so that cgi files can be executed outside any default directories. Edit your conf file so that the following is present:


Options +ExecCGI



Also edit the apache2/mods-enabled/mime.conf file and add:

AddHandler cgi-script .cgi


Test out that a cgi can be executable by creating the cliche hello world script.

2. Put phpshield.5.2.lin on this directory /usr/lib/php5/20060613+lfs so the result will be
/usr/lib/php5/20060613+lfs/phpshield.5.2.lin. This directory is found easily by writing an info.php script




3. Edit /etc/php5/apache2/php.ini and add the following:

extension=phpshield.5.2.lin


4. Reboot apache2.

These are additional steps to the steps listed by PHPMotion here. You may need an account to access this page and I definitely think you need one to download PHPMotion

The New York Times mentioned that a new search engine was being unveiled today: Cuil (which unfortunately stands for “cool”). Looks decent enough. Entry page is clean. Results aren’t bad. I’ve been comparing it to other search engine results and haven’t seen much of a difference other than layout on the results page. You have the option of either two or three column results. The results are also in paragraph form so they’re a little easier to read. I guess it reminds me of the search results layout I’d expect to find when searching online retail sites.

There are some nice tabbed results at the top that break the results up into categories. The one item I wish it had was a way to go to page 2 through n along the top like it has along the bottom since the result page is more like a page and less like a list. For some reason my mouse prefers to stay near the top of the screen when I don’t find results of value on the current page. That way I could click on the results tabs and then go to page 2 or 3 without moving my mouse across the screen. Saving those precious nanoseconds in the name of efficiency you know?

The images that are served up with the results are not from that particular result site but are instead served up from the cuil image servers. Wonder what type of algorithm they use to match the google image below with the search results with it. The search was for “php”.

I put in three weeks notice at the job I’m leaving. I attempted to put in 4 but to do so I would have had to bring in an employee who was enjoying a well deserved vacation. So I now have two weeks left. It’s going to fly by. I’ve been working diligently to prepare the office and my replacement for my departure. I think they’ll be well prepared to coast for a time. I’m feeling a bit like a lame duck and am not convinced that they are in a hurry to replace me with anyone. I’m trying to put everything on auto pilot should that be the case.

The person that I replaced at Hammock remained on for two weeks while I got acclimated to my new surroundings. That was really helpful. If I’m not here when the new hire comes on, it’ll help that I’ve documented everything meticulously (that might be why they’re taking their sweet time finding a replacement – while I’m sweating bullets for them). I think he or she will be able to pick up the loose ends quickly.

The position that I’m taking at another firm isn’t affording me the same starting point. Their IT Manager basically went AWOL. Hasn’t left much in regards to documentation and will have been gone for three weeks before I finally come on board. From the little that I’ve been able to gather from him, their IT infrastructure is a mess. Might not be his fault, he may have been overloaded. Either way, I’ll have my work cut out for me.

Especially since I’ll be transitioning back into the Windows server environment. I can’t tell if they are really dependent on the many features that Windows 2003 Server supports. It may be an opportunity to bring in Mac OS X server. I’ve enjoyed supporting a mixed computing environment. I hope that I can push that there.

This is my first post where I openly state the name of the company I’ve worked at for the past three years. Two days ago I had to make the most difficult decision I’ve made during my professional career. I handed my resignation to people who I admire and respect dearly.

At times during that meeting I felt emotional and at a loss for words, and overwhelmed because I had decided to move on from a company that I very much enjoy working for. Hammock Publishing is an outstanding place to work. There’s not an individual there that I don’t feel has made me a better person for getting to know and work with. And while I’m not a millionaire, I’ve been compensated appropriately for my work. And I’ve had very little oversight, which has allowed me to produce better work. They have been encouraging, understanding, forward thinking, and outstanding people.

I’m not leaving because in their eyes it’s time for me to go. Rather, I decided that I needed to stretch my legs a bit and journey again into the unknown. There is so, so much that I am going to miss. All of that aside, I’ll miss working directly for Rex most of all. He has been like a father and brother to me. I imagine I’ll never work for a better employer.

Thanks.

I feel silly. I support about 25 users who all recently started using Apple Mail so we could leverage our iCal server. Everything had been running fine until three weeks ago. Then one day, we had some DNS issues (I’m really tired of that excuse) and for some unknown reason two user’s Mail programs could no longer send email but could receive it just fine. I chalked it up to the DNS issue and a corrupt setting somewhere and told them to use Thunderbird for a while and moved on to bigger fish.

We suffered the same DNS issues over the past couple of days and low and behold 5 new users suffered the same effects of being able to receive email but not send. Some of you are probably already saying “Duh, did you check…. ” but in my defense I didn’t even think of that since I was overwhelmed with DNS at the time. This morning I started looking through the forums and found that there were a whole lot of folks suffering the same problem and how it just appeared out of nowhere after working perfectly fine.

Cut to the chase, here’s the solution: You’re outgoing email server setting have been changed. Seriously, that’s it. Even thought they were correct, some software bug has decided to set them to default. I’ve seen it happen to nine 7 separate machines at this point. All were running 10.5.2 and the Mail 3.2. So I overlooked what even a newbie should know. Check those server settings closely even if you know you set them correct in the beginning.

Here’s the cause (in theory). If Mail can’t find the outgoing server (ie can’t resolve DNS) there is a very good possibility that it will reset the settings back to non ssl and Authentication to none in advanced outgoing email settings.

Huge break-through today. I was setting up a Sonicwall firewall that I’ve had not end of trouble with. I have ViewPoint 4.1 installed on a windows machine but haven’t been able to get it to actually log data from the SonicWall appliance. I had two issues that may have been related. I eventually realized that I possibly had the wrong info on the appliance for the syslog server. The server IP address was correct but the server port wasn’t set to 514. I set it to that and also enabled all LAN traffic behind the firewall (although I wasn’t seeing anything being blocked).

I also was having trouble configuring scheduled reports on the viewpoint syslog server. In fact, every time I attempted to choose which reports to include, they just disappeared after I pressed Update. Well today that stopped happening and I think it probably had something to do with what I did above. I synced the devices and surprise surprise, it was finally creating beautiful graphs!

I wasn’t able to get this solved using the 245 page admin guide but instead found the solution by following the 7 page guide. That is located in the ViewPoint GUI under Reports > General > Status. You’ll see a link at the bottom of the page title appropriately “Open Getting Started Instruction In New Window”.

We’ve seen an 85% increase in the amount of unsolicited email that is making it to our servers over the past two days. Much of it is the kind that has a forged sender address. We unfortunately have to deal with the spam when it is bounced back from the intended recipients to the innocent bystanders.

All of the emails list:

X-Mailer: Microsoft Outlook Express 6.00.2600.0000


or some other version of Outlook 6. It is beyond me why these folks still have not upgraded their software. Is the IT department even arriving for work anymore?

I’ve also seen these messages from:

X-Mailer: MIME::Lite 3.01


Most of the email pretends to be from (in the body of the html):

aw-confirm@ebay.com
service@chase.com
service@paypal.com


I checked out spam cop and it does appear that there has been an increase in spam on the Internet overall in the past week. I remember reading somewhere recently that it was believed that this was a new version of the Storm Worm Virus. Wish I could remember where I read that. I’ll try to track that link down and post it.

It’s been an exciting morning so far. Within 30 minutes I had these issues:

1) incoming phone calls were all dropped. The tech I called wanted to call me back on my office phone. I had to laugh.

2) several websites were offline due to an apache misconfiguration that oddly didn’t rear it’s ugly head until a week after rebooting the server

3) one of our users had 200 emails in her inbox that were marked as undeliverable and another 198 in quarantine. Obviously she didn’t send 400 emails yesterday. Apparently someone she knows has a virus that is sending out emails with her listed as the sender (sans our email server). Thanks open relay turds.

4) my foot hurts.

I’ve installed 10.5 server several times now and there’s always an interesting issue to deal with. This time it was with the mail server. My original setup and testing of Postfix using the Gui worked well. But eventually something “changed” and the mail system stop delivering email locally.

Here was the error message:

Apr 17 11:19:56 serverhostname postfix/smtp[4678]: DAB0BD6E1E: to=.com>, relay=none, delay=970, delays=970/0.08/0/0, dsn=4.4.1, status=deferred (connect to 127.0.0.1[127.0.0.1]: Connection refused)