Older blog entries for jpalme2000 (starting at number 15)

11 Feb 2002 (updated 24 Dec 2007 at 16:26 UTC) »

Things are better?

Still downloading the updated RPMs for Redhat 7.0. I think I am trailing the rest of the world here by at least .2 revisions. I'm thinking about getting Rawhide, but the problem is that I can find where I can get it.

10 Feb 2002 (updated 24 Dec 2007 at 16:26 UTC) »

Things are fine

Actualy did some integration work today. Had a bit of fun with MRTG and RRDTool, which work as advertised. They'll be fun to integrate into something a bit more interesting.

6 Feb 2002 (updated 24 Dec 2007 at 16:27 UTC) »

Things are Hmmmmmm

Well, another day goes by, wondering what's coming next.

3 Feb 2002 (updated 24 Dec 2007 at 16:27 UTC) »

Things are about the same as yesterday

3 Feb 2002 (updated 24 Dec 2007 at 16:27 UTC) »

Things are ok

VMWare rocks, I can't wait to get a license for it. That way, the ugly fat laptop I have to drag around with me can be used as a REAL system.

24 Nov 2001 (updated 24 Nov 2001 at 17:41 UTC) »

Things are surprisingly good.

Been terribly busy with life since 9/11. It's interesting to read of my postings of 9/10, it shows the innocence of the age of Internet. A time we may look back on and wonder how we could become so relaxed.

My life was "Shaken, not stirred" since 9/11. Busier than a little bee, working on piles of activities. At least I dont' have marketing folks to make happy. Work keeps rolling in.

Been thinking about FastNMS again, I think it's something that I need to get out. I was thinking that a combination of many of the projects out there would help get things started. I really like how Playin' in the LAN approached the design. Perhaps that's a good starting point.

Be well citizen.

12 Sep 2001 (updated 24 Dec 2007 at 16:28 UTC) »

Things are very sad

I'm very sad for those thousands of dead people in New York and Washington. I'm sorry to see all this.

10 Sep 2001 (updated 10 Sep 2001 at 22:41 UTC) »

Things are: Strangely content

Snored my way across the country, enjoyed a nice day of napping on the plane and dreaming about Fast NMS. Feel the design beinging to gel in my head. Feels like the right thing to do.

Maybe I need more sleep, and the lack of sleep is making me depressed? I feel surprisingly good after the naps on the plane today.

Why does everyone like short diary entries?

West Nile Virus? What's up with that? Hope it's not bad.

Have a great day.

Things are: Angst ridden

Here I am, 12:30 in the AM with a 4:30am wakeup for a 8:00am flight to DC. I oh so don't want to go, I wish I was just a happy techie guy again. Nothing better than troubleshooting a T-1 outage at 2:30am, cold pizza and warm coke at your side.

Now, I sit and worry if I can communicate a revolutionary concept to an audience of marketeers, who are much more interested in getting a couple rounds of golf. They'll be sitting impatiently in an dark air conditioned room trying to stay awake for my inept babbling and confusing charts.

So many very cool ideas that will be reality soon, just not my version of them. Unfortunately, I'm 12-24 months ahead of the IT industry, and trying to sell tomorrow's ideas to marketing folks selling to yesterday's clients is both a disheartening and frustrating job.

I only hope that they don't "eat the seed corn" and wack me off the team. We've got some cool stuff, stuff people are going to be screaming for and MS is going to be charging Big Bucks for. Stuff that will change network management as it's known today.

But... that doesn't help me sell it to yesterday's clients. Oh well...

Anyone need a really forward thinking engineer? I'm cheap and carry a wellspring of ability and knowledge.


9 Sep 2001 (updated 9 Sep 2001 at 23:29 UTC) »

Things are: Worse

Thanks for writing everyone, it was really cool to get your notes. And I did totally freak out. ;) (EEK! PEOPLE ACTUALLY READ THIS!) The Things are getting tag was a request from a young man who asked for me to do it.

As for me, thunk up a new/old idea on how to approach network management. Why not make it simple and fast? Use shell scripts and simple command line ping, snmpget, and admin commands to create text files of what is up and what is down. Use a simple gui text reader to read the text files on a periodic basis. Use shell scripts to page you when a device changes status.

By itself, it's not particularly interesting. But if you cut the number of systems you are trying to manage down, and increase the total number of active management shells, you could scale up very nicely. So, let's say taking 254 (a class C for you old guys, a 24bit mask for you VLSM folks) nodes that are being checked and managed. It's called a Management Unit (MU), and a single cron job to run the scripts associated with the MU.

Now, go ahead and have 50 or 100 MUs and run them at the same time on the same machine. Should work just fine. If something changes, write a entry to the Syslog file. Have a nice browser window for checking the Syslog file visually, and a cron job with a script that pages you when something critical happens. (node down, web server offline, coffee break during meeting, whatever).

With this, you could easily mimic the functionality of HP, CA, Netview, or even OpenNMS, and NO JAVA involved. (I like OpenNMS, but I hate the fact they wrote it in Java. Java libraries are buggy and no matter how many times they tell me I suck, I'm not going to change my opinion about how terrible the libraries are.)

So, a Fast, Simple, easy to implement NMS system that any Junior System Admin can figure out and actually run.

YOu might wonder about the discovery portion. That's pretty easy too, many tools exist to do this. For example; Use NMAP to discover the network and identify the OS, the services running on it (SNMP, Web, DMI, whatever). Then take the output of this and pipe it into a script that calls other scripts to configure the proper management tools for it.

So, let's say that nmap gives you a really good return on a machine. It would look something like this... Starting nmap V. 2.54BETA7 ( www.insecure.org/nmap/ ) Host localhost.localdomain ( appears to be up ... good.

Interesting ports on localhost.localdomain ( (The 1523 ports scanned but not shown below are in state: closed)

Port State Service

21/tcp open ftp

22/tcp open ssh

25/tcp open smtp

80/tcp open http

111/tcp open sunrpc

199/tcp open smux

443/tcp open https

1024/tcp open kdm

3306/tcp open mysql

6000/tcp open X11

8080/tcp open http-proxy

Nmap run completed -- 1 IP address (1 host up) scanned in 3 seconds

From this, we would know that we could run SNMP tools like RRDTOOL to managed services such as ftp, sntp, http, https, mysql, X11, etc.... Other scripts could automatically configure the proper templates to start the management of these devices.

The best part, it's all VERY EASY to ready and there are no dumb databases, Java, SQL, or other crap in the way. Just you, your shell, and your script files.

Since everything would be text file based, it would be easy to integrate other tools into the set making a really cool set of thingies to manage anything that can be reached via a command line.

What more can you ask???

--<Edit from the future>-- Doubleplus good citizen.

6 older entries...

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!