http://www.randombugs.com Just another Bastard Operator From Hell. Everything from *nix to programming Thu, 06 May 2010 14:21:46 +0000 en hourly 1 http://wordpress.org/?v=3.0 http://www.randombugs.com/java/glassfish/changing-logging-system-glassfish-21.html http://www.randombugs.com/java/glassfish/changing-logging-system-glassfish-21.html#comments Mon, 28 Sep 2009 18:04:20 +0000 admin http://www.randombugs.com/?p=606
A lot of people are not satisfied by the default Java Logging (JUL) in Glassfish. Also, myself I encountered some problems on clustered environments where we should have one single log and not one log for each instance. Trying to change the logging system in Glassfish it looks to be almost impossible , but I found a very interesting project java.util.logging to log4j Bridge and the sky become more clear for Glassfish logging. Jul to Log4J Bridge are part of log4J project and is very useful when you ...]]> http://www.randombugs.com/java/glassfish/changing-logging-system-glassfish-21.html/feed 0 http://www.randombugs.com/linux/tuning-linux-system-database.html http://www.randombugs.com/linux/tuning-linux-system-database.html#comments Thu, 16 Jul 2009 17:24:13 +0000 admin http://www.randombugs.com/?p=567
In my experiments with databases I was constrained to tune the system together with the databases. A system what will run a database is quite different from any other server system, because databases put a lot of stress on the IO and especially on the disks (probably the file servers too). Hardware is not all you should tune your system as much as possible. In this post I will talk about Linux with Ext3 (is what I have right now), but on future I will test OpenSolaris with ZFS. 1. ...]]> http://www.randombugs.com/linux/tuning-linux-system-database.html/feed 2 http://www.randombugs.com/tuning/compile-install-run-modlimitipconn-apache2-debian.html http://www.randombugs.com/tuning/compile-install-run-modlimitipconn-apache2-debian.html#comments Mon, 22 Jun 2009 17:17:48 +0000 admin http://www.randombugs.com/?p=509
Limiting abusers from running 20 download threads from the same Apache server is easy now with mod_limitipconn. But the apache 2 module, mod_limitipconn, is not available under the latest Debian distribution, Lenny. In this mini how-to I will show you how-to manage this module to work for Debian Lenny users. The compile process of mod_limitipconn is a little bit more than “make install”. First install apache prefork development utilities (we need apxs2 to compile our module) [rb@randombugs]# apt-get install apache2-prefork-dev After that, download source code of iplimit module from http://dominia.org/djao/limitipconn2.html and unpack it [rb@randombugs]# ...]]> http://www.randombugs.com/tuning/compile-install-run-modlimitipconn-apache2-debian.html/feed 3 http://www.randombugs.com/linux/mysql-postgresql-benchmarks.html http://www.randombugs.com/linux/mysql-postgresql-benchmarks.html#comments Thu, 18 Jun 2009 14:25:25 +0000 admin http://www.randombugs.com/?p=458
We are living interesting times ... MySQL was first purchased by SUN and now SUN was purchased by ORACLE. I don't know what future will reserve for MySQL, but in this moment it seems MySQL is coming very very close to PostgreSQL. We are NOT talking about MySQL with MyISAM, we are talking about MySQL with InnoDB, because I'm directly interested in a set of properties what PostgreSQL already have them built-in and MySQL achieve them through InnoDB (and the new Maria Data plugin). This properties are Atomicity, Consistency, Isolation, Durability = ACID, in other words, very stable, good integrity and crash proof database. Why an ACID database? Sometimes we are more interested in ACID for our data than raw speed. For example do you keep your savings to a bank who is running a NON ACID database? I think you understand my concern. ]]> http://www.randombugs.com/linux/mysql-postgresql-benchmarks.html/feed 45 http://www.randombugs.com/linux/gpg-ubuntu-debian.html http://www.randombugs.com/linux/gpg-ubuntu-debian.html#comments Wed, 13 May 2009 12:40:28 +0000 admin http://www.randombugs.com/?p=443
Gpg is stands for Gnu Privacy Guard and is a free alternative to the PGP cryptographic software. GnuPG (or GPG) follows the RFC 4880 which is the standard specification of OpenPGP. The most important thing GPG is interoperable with PGP. GPG is build as a command line utility but also have several front-ends for KDE, Gnome and other Linux desktops, but also is directly integrated in other software like: Email Clients as Mozilla Thunderbird, Evolution, Kmail, Instant Messaging as PSI, Fire, Browsers as Mozilla Firefox etc. GnuPG encrypts messages using ...]]> http://www.randombugs.com/linux/gpg-ubuntu-debian.html/feed 2 http://www.randombugs.com/linux/migrating-ext3-ext4.html http://www.randombugs.com/linux/migrating-ext3-ext4.html#comments Tue, 24 Mar 2009 07:53:49 +0000 admin http://www.randombugs.com/?p=347
Ext4 is the normal evolution of the most used Linux filesystem: Ext3. The evolution of Ext3 from Ext2 had add just 1 new important feature: journalization. But ext4 is a result of hardly improved Ext3 and is coming with better performance, modified data structures, reliability and new features like bigger filesystem/file sizes, subdirectory scalability, extents, multiblock allocation, delayed allocation, journal checksumming, online defragmentation, persistent preallocation. The first stable version of ext4 is already out and probably a lot of you want to test it a little bit, so I write ...]]> http://www.randombugs.com/linux/migrating-ext3-ext4.html/feed 0 http://www.randombugs.com/linux/patching-linux-kernel-rebooting.html http://www.randombugs.com/linux/patching-linux-kernel-rebooting.html#comments Thu, 12 Mar 2009 15:23:45 +0000 admin http://www.randombugs.com/?p=268
I think, everyone from Linux industry, knows about kexec (fast rebooting ... actually not rebooting just switching kernels) or Pannus (live kernel patching, but right now the project is dead) and their use. The Linux community started very earlier to think at a method to switch or patch a live kernel, but this things should work only on a very good kernel with very good memory and process isolation ... but this was achieved some time ago, we have a very good and mature Linux kernel (in my opinion was perfect from 2.4.2x) what can run for years without a reboot (my record is 4 years on a system with kernel 2.6.8 and this kernel is far to be perfect). This days I found a tool what is a gold mine for system administrators : KSplice. ]]> http://www.randombugs.com/linux/patching-linux-kernel-rebooting.html/feed 0 http://www.randombugs.com/linux/converting-ext2-ext3-ext2.html http://www.randombugs.com/linux/converting-ext2-ext3-ext2.html#comments Wed, 04 Mar 2009 14:16:41 +0000 admin http://www.randombugs.com/?p=210
Yes, is possible and is very easy. Ext2 differ from Ext3 only on few things: Journalling is one of them and can be added very easy to an Ext2 system. First, just look to the features of a Ext2 filesystem: [root@randombugs]# tune2fs -l /dev/sdb1 | grep features Filesystem features: ext_attr resize_inode dir_index filetype sparse_super large_file Adding a journal to a Ext2 filesystem is an easy process and some people are doing that directly on a mounted partitions. I recommend you to don’t do it on a read-write partition ...]]> http://www.randombugs.com/linux/converting-ext2-ext3-ext2.html/feed 2 http://www.randombugs.com/linux/rebooting-kernel-panic.html http://www.randombugs.com/linux/rebooting-kernel-panic.html#comments Wed, 04 Mar 2009 13:24:54 +0000 admin http://www.randombugs.com/?p=205
By default, a Linux, is waiting direct input of a person / sysadmin when is crashing with kernel panic/oops. Obviously is very important to know or to see directly the dumped screen, but sometimes in production environments is better just to reboot itself without any intervention and debug the problem with the system online. Off course for debugging and seeing the dump you will need to install and configure kdump, but that is another story (I will probably write about that soon). Anyway to set your system to reboot after 1 ...]]> http://www.randombugs.com/linux/rebooting-kernel-panic.html/feed 0 http://www.randombugs.com/linux/core-dumps-linux.html http://www.randombugs.com/linux/core-dumps-linux.html#comments Fri, 27 Feb 2009 12:47:55 +0000 admin http://www.randombugs.com/?p=194
If you reached this page is possible to know what a core dump is, but I will explain again (in a very short version) for the new “kids” on the Linux who also is possible to reach this page … it sounds like a paradox … whatever: Core dump is a file what contain the current state of a process and its memory saved when a program crashes. It can be analysed with gdb but you need programming skills to understand and develop that, Core Dumps are very useful in a development ...]]> http://www.randombugs.com/linux/core-dumps-linux.html/feed 1