Tag: self-hosting
-
Cold Storage: From Manual to Automatic Backups
My files live on two Synology NAS units. The DS920+ is the one everything points at. The DS1621+ holds a second copy, kept in step by Synology Drive ShareSync. A change on one shows up on the other within a few minutes.
That’s good protection against a dead drive or a dead NAS. It’s no protection against me. ShareSync is a two-way sync, so a deleted folder, a bad script or ransomware reaches the second NAS just as fast as a good change does.
Read more about Cold Storage: From Manual to Automatic Backups
-
What my homelab costs to run: 337 W and S$76 a month
I priced my homelab in a spreadsheet, then checked each row against what was actually running.
The homelab’s devices draw about 337 W across 21 rows. After UPS losses that’s about 377 W at the wall, or 276 kWh a month. My rate is 25.39 c/kWh before GST, 27.68 c with 9% GST, so the homelab costs S$76.25 a month, about S$915 a year.
Read more about What my homelab costs to run: 337 W and S$76 a month
-
Homelab hardening: what, where and how to start
You don’t need a security background to make your homelab a much harder target. You need an order of operations.
I spent an afternoon getting Claude to scan my homelab properly for the first time. I expected the scary findings to be at the edge: an open port, a forgotten forward. The edge was fine. The problems were all inside, in the boring places I’d never look at since things worked.
Read more about Homelab hardening: what, where and how to start
-
Fourteen Lists, One Job
I replaced fourteen overlapping Pi-hole blocklists with one maintained list, but only after replaying a week of my own DNS traffic against each candidate.
A Reddit post made the case for merging every blocklist you use into one “master” list, hosting it on GitHub, and pointing Pi-hole at that instead.
-
The Detector Was Never Looking
A neural network running at 5.8 milliseconds an inference, pointed at a room with cats in it, that had never once detected a cat.
I have an NVR. It runs in a container on one of the cluster nodes, it does object detection on that node’s integrated GPU, and it is fast: 5.84 milliseconds per inference, sitting at about 6% of what the detector can do. Person events come in with confidence scores between 0.73 and 0.93, which is healthy.