Slashdot Mirror
CentOS Project Administrator Goes AWOL
An anonymous reader writes "Lance Davis, the main project administrator for CentOS, a popular free 'rebuild' of Red Hat's Enterprise Linux, appears to have gone AWOL. In an open letter from his fellow CentOS developers, they describe the precarious situation the project has been put in. There have been attempts to contact him for some time now, as he's the sole administrator for the centos.org domain, the IRC channels, and apparently, CentOS funds. One can only hope that Lance gets in contact with them and gets things sorted out."
This kind of thing really scares me, because this is exactly what it happens when someone dies, for example: the data/information stream coming from them on the web simply ceases to exist. Also, this is one of the main reasons why important projects should have their main assets handled by a group of people, and not have things centralized. If the worst has happened, CentOS will be forced to fork their project and start over.
But let's just hope I'm spewing bullshit and he's just pissed off.
Somewhat concerning, considering the number of CentOS servers I have in the wild.
I'd suggest disabling yum updates on your CentOS boxes until this gets sorted out. Might want to do updates by rebuilding src rpms directly from Redhat.
Just the fact they even have to address an issue like this makes me nervous.
Another consultant who stuck it out.
"We are the Priests, of the Temples of Syrinx..."
Ubuntu Server?
No offense to the Ubuntu team intended (or to you) but that's not exactly a hardened OS with the kind of long term support one needs in a data center.
If low budget to you is a simple LAMP stack- then maybe. But no one has been beating up on Ubuntu server- and it really needs professional QA before anyone tries to use it for more than a novelty.
The logical alternative for new deployments would be Debian, if you wanted to dump RPM based systems.
Another consultant who stuck it out.
"We are the Priests, of the Temples of Syrinx..."
Ubuntu has slowly made it's way to the data center over the last couple years and it's doing quite well. Typically admins will use the LTS versions which are supported for 5 years. You can also upgrade directly from one LTS version to the next LTS when it comes out, no need to hit any of the minor version in between.
Ubuntu is seeing HEAVY use in virtualized environments, like Amazon EC2, and since it's built off of Debian it inherits much of that distribution's stability and polish.
I've been a professional Linux admin for 15 years, have run everything from Red Hat, Cent OS, Gentoo to Debian in the data center and definitely think Ubuntu Server has its spot in the data center as well.
I totally agree with the "bus syndrome" thinking. At one company I was at for many years, the boss had that concern. I was the senior SysAdmin, who created and managed the entire IT infrastructure. To alleviate it, everything was documented. A copy of the passwords were kept under lock and key. Server functions were well documented. My assistant(s) (depending on the year I had 1 to 4) could continue smooth operations without me.
Keeping the "bus syndrome" mentality, should I be unavailable for a day or days, there were no problems. I could fly between cities to do work, and not panic that the whole world was going to fall apart while I was on a plane. I still got plenty of phone calls, simply because it was my baby. Junior admins didn't want to make widespread changes without my seal of approval, even if it was a quick phone call where they gave me a brief outline of their changes, and I gave them verbal approval ("Go for it. Let me know how it goes.").
The day came that they decided I wasn't necessary. I was locked out of the machines per my own plan, and then notified that I was no longer part of the company. Whoever did the changing wasn't quite as consistent as I was and missed a few spots. Being a "good guy", I verified that I was locked out of everywhere, and sent a list (it was short) of what I still had access to, so they could get those too. The missed spots were non-essential, so even if I had a desire to do bad things, I couldn't have broken much.
The password plan had better motives than firing the top guy. On password change day, I issued the passwords on slips of paper to the people who needed them (and to the vault). Should someone's passwords become compromised, I could have all the passwords changed in approx 5 minutes. Should something seem funny, we'd change the passwords. Usually we just changed them because the existing passwords had been in use for too long. We did have someone lose their USB key with their SSH keys on it. We went through the well practiced drill. It turned out to be just an exercise. The key had fallen out of his pocket, and was under the seat of his car.
When they terminated me, the company lived on. The transition was smooth without me. I may as well have been hit by a bus. No one asked me "how do I....?", because it was all there for those with access who knew what to look for. Even if we had a walkout of all IT staff, things were documented well enough where an experienced IT person could walk in and keep things running.
We were a high dollar, small staff company. Why should somewhere like CentOS be any different?
Serious? Seriousness is well above my pay grade.