Building and compiling

Nothing much today:

  • crossover plugin finally started to work (problem was with the NIC, not with the plugin itself).
  • MySQL rpms for version 3.23.49 built good, but no changes noticed in handling foreign keys/references. Maybe it’s just me.
  • 2.4.18 kernel rpms were built good too. Probably I will upgrade my machine tomorrow. Reasons for upgrade: new scheduler and fixes in USB.

Random bits

While studing the subject of migration from Microsoft to Linux, I came accross nice description of how to set up authentification of Linux box in NT domain.

I also changed all links on this page in such a way that printed version is readable and actually usefull :)

I have installed bash programmable completion. While it is a littl bit noisy about me not having bash of exactly the version it needs, I still get a lot of satisfaction. My TAB hit/miss ratio has grown incredeably, since I was pushing TAB key for most of the times I use it now anyway (huh? :) ).

Officially, from today I am using Mozilla instead of Opera. Opera is fine browser and it suits all my needs, but it is not an open source software, which I wish it very much to be. Idiology…

New mutt packages, Quake III, and book order

I have put online new mutt rpms and configs with S/MIME support and LDAP address book configuration. No NNTP and compressed mailboxes support this time. Outlook appointments converter to ical calendar is available though.

After long tweaking and changing all possible and impossible settings I actually managed to fix my Quake 3 and now I get up to 140 frames per second. Thanks to [bofh]-mokh and [FC]-Pekker for their help.

I have also decided that it is time refresh my home library. I have ordered several books from O’Reilly. This time I tried not to limit myself with one topic and asked for books on everything :) (vi, cvs, php+mysql web programming, exim, perl for system administration, etc). Waiting is painful!

Linux HA clustering, mutt with LDAP, and workstation upgrades

Last few days I have spent installing a HA (High-availability) cluster. I needed an active/active setup for two nodes to run Apache and PostgreSQL services. Both nodes are connected to external shared storage. It took me some time to get the concepts (especially with quorum partitions and raw devices), but afterwards, installation was a bliss ;). I have used KimberLite clustering technology, which is open source and free. There is also a commercial solution on the site, for those who want. SourceForge.NET Clustering Foundry was also very helpful in getting the whole thing.

Meanwhile, I managed to configure my mutt to work with MS Exchange address book over LDAP. mutt-ldap-search.

I’ve also had my workstation upgraded in certain ways :) More memory, less CPU, better graphics card (GeForce 2 MX 32MB 200 Series this time). Strangely, I cannot get more then 40FPS in Quake3 using original Nvidia drivers.

Now Last of all the news is that I am now testing init script for NovaNet services. Since Novastor didn’t provide with anything automatic to stop the service and remove all kinds of shared memory segments and semaphores, I had to write my own. It will be available for public, when initial testing is over ;) .

Kickstart

Recently I was playing with KickStart – this piece of software helps you in case you need to do plenty identical installations. ksconfig is a new configuraiton tool, which helps you to create installation config file, which can later be put on the floppy for UNinteractive installation :). I used it for cluster setup where nodes are pretty much the same with all the difference set in the DHCP server.

CVS and bugtracking with Mantis

All the small things to close working week. Rest of the day I’ve spent education myself on CVS again. :) I’ve came across CVS book at red-bean.com which I now want to order hardcopy. All my today CVS practise was around a wonderful piece of software called Mantis. It is simple yet powerfull bugtracking system, which I have adopted as technical support tool for our company. All of my changes were based on version 0.16.0, while mantis went to version 0.17.0. So, I’ve been porting all my changes to the new version. Tried both ways with traditional diff-edit-patch-repeat and cvs ways. So far, I enjoy the cvs way of doing things, and I haven’t yet tried branching and merging. Oh, I sound like a developer, don’t I? :)