I have been chasing after squirrels lately… lots and lots of squirrels. I recently proposed “should I build my own distro [for the company]”. I was thinking about the potential for a tech pandemic. I also started thinking about machine learning and processing of syslog messages from systems and applications.
I hate aggregated logging because at some point you have to decide between verbosity, performance, and capacity.
A few months ago I tinkered with Synology’s VM tools. The problem is that the Synology does not have enough CPU or RAM to do anything meaningful. Even though they have apps and the system is mostly idle… it does not take much to redline.
At the same time I have been using my Synology as a VPN server. They has been working ok but recently it started dumping my connections every so often. I have been tracking the logs to see if there is something identifiable going on and it’s still not clear.
But then I look at the logs and I see all sorts of crazy unidentifiable messages. So when I look at other enterprise installations. Do those operators really know what’s going on in their datacenter? And so I’m back to the consideration… should I build my own distro?
docker is an easy way to assemble a distro but it’s still less than ideal.