Known Element Enterprises Corporate Blog

I created a twitter account today. Often I will want to post something online that I am doing, thinking about, viewing etc. This blog didn’t quite seem like the place to do it. I try to reserve it for longer articles and thoughts on various topics. Although I have at times posted quote or links, and seemingly random things (which meant something to some people, a puzzle if you will). 

My twitter account can be viewed at:
http://twitter.com/charlesnw

As I go throughout the day, I will post things of interest or updates on what I am working on. I am not going to post every minute detail (like when i am eating meals or going to bed etc). However as I work through various problems (setting up Groundworks for instance) I will probably post quick snippet updates on my twitter page. I can post from my phone via txt message and from my AOL IM Account (I refuse to call it AIM). 

I also created a facebook account and joined a few technology related groups and the Los Angeles Network. I want to use Facebook and the group functionality it offers for LUG related things (such as hosting Photos/Videos/Podcast snippets). Also I am looking into ways to integrate the calendar with WebCalendar and provide a central place for SoCal open source events. 

So I have very specific uses planned for my foray into Web 2.0, mostly as an extension of things that happen “in real life”. I would say that 90% or more of the people I interact with online (via e-mail and irc) I also see at least once a week or talk on the phone with them. I am not really trying to make new friends online per se. In person human interaction is very important. The interwebs can never replace that.

I have been running LiLAX for a while now and thought I should write about what it takes to run a user group. 

First of all I schedule topics/speakers in 6 month blocks. I put an initial topic list online, and discussed it at the first meeting of LiLAX I ran. The members wanted more desktop related stuff, whereas my list was mostly focused around server oriented topics. So I changed the line up, and last month held part 1 of the Open Source Desktop Bakeoff. This month will conclude that topic. 

The key thing I learned from that, is you need to put an initial prototype for your users to revise. If you simply ask people what they want to see you won’t get very far. I had a particular type or category of things I thought the group might be interested in and they let me know otherwise. Which is totally cool. Obviously if I don’t cater to the needs of the group members, then they will be unhappy. On the other hand, they need to be gudided and have some structure to operate in. Otherwise the group will not move forward.  I guess you could say the governance model is a democracy where I hold veto rights and hopefully never have to exercise them  

The second thing I do is record the presentations I attend and post the recordings online. I make sure to do it for LiLAX (although my recorder died during the last meeting), and I try and provide that service for other groups as well. The service is still in a bit of beta and has some rough edges to be worked out but is coming along fairly nicely. In the future I would like to do video recordings and eventually streaming. This is something I will look into next year. 

The thirdthing is being organized and contacting speakers early and often. The reason that the desktop bakeoff was such a success, was that I contacted all the involved parties well in advance and worked on creating an agenda and what I would like to see covered. I think a lot of presenters may have canned topics/slides, but are looking for some gudiance.  

Thats about it so far. Also I have setup reminders on my calendar to send out announcement and wrap up e-mails and post the podcast. Otherwise I will forget

requires plenty of sugar and caffeine:

http://bp0.blogger.com/_NJ4JFBfr1tY/Rs3objhDtzI/AAAAAAAAALQ/Ls4xKhA-F28/s1600-h/energy.JPG

http://hey-genius.com/ 

 While reading over my RSS feeds today, I came across the following two links:

http://www.vmware.com/community/thread.jspa?threadID=86877http://www.vmware.com/community/thread.jspa?threadID=94743

both of them providing much detail on how to setup ESX in Workstation 6.

Very cool stuff!

Recently I requested an evaulation of ESX server. I received the license keys and download links and started reading over the documentation. 

The first step was to install the license server, virtual center and msde. No problem. I put those on my Windows 2003 Guest (along with VmWare Server client package).  Easy as pie. The only snag I ran into was selecting the wrong license file (VmWare sent me two). Reading the licensing chapter in the VmWare install guide made my head hurt

Second step was installing ESX Server. Reading through the requirements I don’t see anything requiring the use of physical hardware. 

*rubs hands with glee* 

I installed ESX into a VmWare Server guest and it works quite well. Evidently other people have done this as well.

I was very unimpressed with Hyperic. The install failed. The agent wouldn’t start due to bugs in the startup script. I give it a failing grade. I will now be installing Microsoft System Center essentials 2007 and evaluating that.

Try again Hyperic.

My previous attempts to deploy monitoring solutions involved Nagios and Nagios based products (like Groundworks). I also deployed Zenoss and Zabbix.

Nagios has always worked very well for me, but requires a fairly substantial amount of effort to deploy.

Groundworks was fairly nice and I may evaluate that some more.

Zenoss and Zabbix looked very nice, but I had a hard time understanding the products user interfaces.

A new product which I recently discovered is called Hyperic. It won best of product for systems management at LinuxWorld this year, and was recently mentioned in the port25 blog at Microsoft. I will be evaluating the product today.

The second stage of my monitoring project is determining what I want to accomplish.  I do that by asking why I determined a monitoring project was needed in the first place:

1) I have a dynamic IP address on the server that hosts various personal and organizational domains. I need the ability to determine when it has changed and alert me so I can update my DNS records. 

2) I have experienced some out of memory errors and service crashes that have gone undetected. I want to be alerted to service failures, and have corrective action taken. That could range from a service restart to complete server fail over. 

3) How well are my systems performing? Do they have enough memory and hard drive space? (See out of memory errors mentioned above). I need to perform capacity planning exercises and having performance data avaialble makes this much easier. 

4) I have a number of  VmWare images running on my server. VmWare Server has several performance issues when you get up over 2 or 3 virtual machines on one physical system. I have made several changes to my VmWare settings that have resulted in improved performance. However my methodology for determining the improved performance was fairly ad hoc. As I continue to make changes to both VmWare settings (global and per guest configuration) as well as inside the guest (to kernel and userland) I want to see the impact the changes have. 

You could call the above items a business case for justifying  expenditures of resources. A problem (or series/category) of problems was identified and summarized. 

So the first item on my TODO list for the remainder of the year is a monitoring solution.  I have covered this before, now its time to sit down and actually implement a solution  

The first thing to discuss is the environment I am monitoring
1)  Personal servers at my house (hosting media/various websites/files etc). 
2)  Servers I administer outside my house (e-mail/web for thewybles.com, my xen host, lilax.net host, client websites) 

In that environment what do I want to monitor?

1) Availability. I have a number of different systems and applications deployed in my environment:

A) RDP/SSH 
B) Apache/IIS
C) Postfix
D) Exchange/IMAP
E) Active Directory 
F) MySQL/MSSQL/Oracle 

I want to make sure the services and applications are avaialable and if not send me an alert. I need to have various levels of checks performed ranging from simple port connectivity, to page content matching. 

2) Performance and load. How well is the system performing? How much memory is being used by the application? What is the system load? How much network bandwidth is being used?  And of course the normal system health checks like overall CPU/Disk/Memory usage. 

3) File integrity. Are critical system files free of unauthorized modification. 

4) Network integrity/intrusion detection and prevention.