Food and water getting scarce!!
Well this is week 7 since I've been layed off. The job situation has been kinda bleak. There seems to be more jobs now that it's after the holidays but I still dont have one. This is the second time in 2 years that I have been layed off.
Is there any hope that the economy will turn around?
I've been working on a few different things.
MySQL cluster: Brian Aker (krow) and I have been discussing different ways to make a MySQL cluster. All of the known ways seem to have faults. I went ahead and made one that uses PVFS. What I did was make one master/metadata server and 3 IO servers. PVFS distributes a files data between the IO servers. The master keeps track of the meta info and where what parts of the file are on what server. I built MySQL on the master server and made the MySQL database files sit on the PVFS filesystem. Unfortunatly the benchmark results were very poor. So right now I'm playing around with the MySQL code to maybe have it use the PVFS library. This might help with the speed of it. There are a couple of other things I might try to see what speed increase I get.
Honeynet Project: The Linux kernel patches I put together were mentioned in the honeynet projects white papers "Know your enemy: Honeynets". This was a pleasant surprise.
openMosix: I havent done a whole lot for openMosix these last few days. But I have some plans in mind about openMosix's filesystem oMFS. I'm still thinking about it though so nothing I want to make public yet.
Well.. thats it.
I've been working some on the openMosix project. I did some kernel cleanup recently that should be applied shortly. I also have some more cleanup patches to send in soon. Nothing big. Mainly trying to get the openMosix code cleaned up so that we can get it included into the linux kernel. Hopefully that will happen come the 2.7 kernel series. We shall see....
I've also been hacking on some kernel patches for the honeynet project. The patches impliment command logging to syslog or a file. You can get them here for anyone interested in them. I also worked on a terminal patch for the linux kernel that logs all terminal input and output to a file. But so far it doesnt look very feasable to do. It seems to log everything.... including term settings. So for now it's going on the back burner.
Well thats about it for now.
Decided to rewrite the openMosix userland code. The current version is based off a old Mosix version that was GPL'd. But we really need our own if you ask me and so far David Santo Orcero, Jeff MacDonald, Rodrigo Henriquez, and myself are going to rewrite it from the ground up. I might have to put it off for a little while though because of some personal things going on in my life, but it shouldnt be too long.
I recently setup LXR for openMosix at this site. It's a very useful tool for learning the openMosix kernel code and also good for developing if anyone is interested in helping.
Thats about it for now.
Been coding alot lately. I started applying some of the kerneljanitors TODO list to the openmosix code. Still have a long way to go though before I really understand it (kernel) all. It's all a matter of time I guess.
Havent been doing much online gaming lately. Dont have the time really. Between work, the family, and my coding projects I dont have much time to play any of the games.
Works been pretty much the same. The normal System Admin stuff.
I havent been coding much lately. I should be working on one
of the many personal projects I have. Instead I have been
playing on line games. What a waste. Oh well...
I started hacking around with the linux kernel recently. I blew one up the other week like you wouldnt believe. People would have been proud. :-) I'll have to set up logging over serial so that I dont have to copy down the oops message. It's a real pain in the ass.
I set up lxr for the project openMosix that I belong to. We just recently announced the porting of openMosix to IA64 (Thanks Matt).
Well thats about it for today.
This weekend I rewrote a function that I wrote over a year ago that makes the random quotes for www.chathouse.com The previous function was based on a fixed amount of quotes in a file and this new one doesnt care about how many quotes are in the file. So you can add and remove them as needed.
I also worked on openMosix userland tools. Particulary the setpe code. I'm not done with it yet but I should be able to submit it soon.
Well thats about it for now.....
I've been doing alot of different things lately. At home I've been hacking around with the openMosix userland code. I would like to hack around with the openMosix kernel patch and code but I dont know the linux kernel yet. I'll learn it sooner or later.
At work I've been messing with drbd (http://www.complang.tuwien.ac.at/reisner/drbd/) and heartbeat (http://www.linux-ha.org/) for a HA Linux NFS server. It's all working well now. I'm stress testing it today. I made a program that copies and deletes files on the drbd partition. Seems pretty stable so far and replication is working. The initial sync is a little slow though. The failover works pretty well too. It takes about 20 seconds or less to failover.
Got my openMosix cluster working finally. I made a foolish error of locking ssh. Everytime I ssh'd to a node and ran a program, proccess migration didnt happen because I locked ssh along with everything else. One needs to unlock it so that his/her proccesses are allowed to migrate to other nodes if need be...live and learn. It's a small cluster but it's enough for me to hack around with the source code and make sure my code actually runs.
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!