Hardware? Yes, “server hardware” I guess you could say. Actually is just a firewall. I bought a bigger firewall a while back and its been sitting here for a while now. Time to finally do something with it. Not a ton of work got done but some did.
Started with the developer meeting. Not a ton to discuss, but we covered some of the AI topics. I talked to them about being able to later integrate with AI agents like Siri or something else. They thought it was a good idea. We may look at filing for a patent for that one as well. Had to upgrade some ram on a development server because we run a scraper on there. We run our own equipment which is good and bad. We save a lot of money by not having to pay AWS or whoever a ton of money for mediocre little servers but we have infrastructure to maintain. Its kind of fun though. I like going down to the datacenter not knowing wtf Im doing and getting our equipment going. Its stressful but rewarding.
So our firewall has been a slight choking point I feel like when we do our big weekly backup to Backblaze. The cores in the firewall are overloaded running at 90-100% for several hours. I noticed this a while back and now its finally time to upgrade the firewall. Its twice the size but nothing like a full datacenter would run. Its way more than enough to host our application, backup servers, NAS, etc. I was happy I was able to just take all of the config from the old one, export it and then load it on the new one. I’ll have to make some changes at the datacenter, so we will see how that goes tomorrow. I figure Thanksgiving is a good day to take the system down for a little planned maintenance. Wish me luck.
Oh, one more thing. I was looking at running these two firewalls for high availability. Have a failover one just in case the new bigger one fails, we are running the old one. I just don’t understand though. Everything always ends up with a single point of failure. Its almost pointless in my mind. I’m better off getting some more server space at another datacenter and just have a full fail over network. If we run two firewalls, they have to connect to a single switch. If that switch dies, it doesnt matter if the firewalls aren’t working. If the upstream Cogent switch, router, whatever dies, the switch doesnt matter if its running. No matter what, there is always a single point of failure. We will plan for a redundant network in the future.
