Sysadmin foundations

The non-negotiable base layer under every specialization — this vault’s owner-statement of what a sysadmin/sysops/devops/SRE must actually understand, independent of their lane.

The take (USER, 2026-10-20 — translated from Polish, original below)

Although I’m not a database specialist, this is exactly the kind of thing worth having a notion of and spending time to understand — solid foundations, regardless of what you’re actually interested in. That’s my private opinion, and I think it’s worth it because 99% of people haven’t the faintest clue, unless they genuinely went down the DBA path. And admins/sysops and devops — and especially developers, programmers, and people right at the boundary, like SREs — should in my opinion have solid foundations in how databases work, how to work with them, and what to pay attention to.

Because no amount of muscle saves a bad job — if the code is crap, it’s crap (and in this case: if the queries are crap, they’re crap). And not every stupid decision can be buried by throwing more RAM or CPU at it. In most cases, EXPLAIN and adding the right indexes is enough.

The second aspect — which is and always will be in demand beyond databases — is networking: solid foundations plus IPsec tunnels, VPNs, VLANs, and BGP. I’m at the stage of foundations + VPN and IPsec. BGP (FRR, VTI, vtysh) is waiting its turn.

Original (PL): chociaż nie jestem specem od baz danych to akurat to jest coś o czym dobrze jest mieć pojęcie i poświęcić czas na zrozumienie i jakieś solidne podstawy bez względu na to czym się interesujesz. No ale to moje prywatne zdanie i uważam że warto bo 99% ludzi nie ma zielonego pojęcia chyba że ktoś faktycznie poszedł ścieżką DBA. A i admini/sysopsi czy devopsi, a już szczególnie deweloperzy, programiści czy ludzie stricte na styku jak SRE moim zdaniem powinni mieć solidne podstawy jak działają i jak pracować z bazami danych czy na co zwracać uwagę. Bo i herkules dupa jak kod to kupa (czy w tym przypadku zapytania do bazy). A nie wszystkie debilne decyzje da się zasypać dosypując więcej RAMu czy CPU. W większości przypadków wystarczy EXPLAIN i dodanie odpowiednich indeksów. No i drugi aspekt który jest i zawsze będzie w cenie poza bazami danych to sieciówka - solidne podstawy + tunele ipsec, vpny, vlany i BGP. Ja jestem na etapie podstaw + vpn i ipsec. BGP (frr, vti, vtysh) czeka na swoją kolej.

Pillar 1 — databases (the vault’s existing lanes)

• PostgreSQL anti-patterns — the top-10 field guide; items 1–5 are literally “EXPLAIN and the right indexes”
• MySQL logical vs physical backups / MySQL backup strategies — the backup/DR dimension of “what to pay attention to”
• Betting on Percona — the packaging decision those foundations inform
• jsystems Kurs administracji PostgreSQL (63 lessons) — the structured deep-dive already captured
• Things I want to dive deeper — EXPLAIN ANALYZE hands-on is the declared first labs session; the anti-patterns note is its syllabus
• The USER maxim maps onto the Sysops interview questions anti-pattern probe #2: “database overloaded — which metrics before scaling?” — the same lesson from the interviewer’s side: measure before you buy hardware

Pillar 2 — networking (the vault’s existing lanes)

• TecMint’s 20 networking questions — the captured anchor canon (Sysops interview questions); Q20-style diagnostic walks are the standard
• Things I want to dive deeper — distsys reading (aphyr) covers the theory side: causality, networks-as-message-flows, when networks go wrong
• Homelabs / pfSense lane — VLANs and VPN/IPsec in production-at-home practice
• BGP (FRR, VTI, vtysh): declared next — not yet started; when it starts, the natural path is FRR on a lab VM + a eBGP session between two namespaces, then the homelab’s MikroTik as the real counterpart

Why these two

Everything else a sysadmin touches (kubernetes, IaC, observability, agents) is downstream of these two: every meaningful workload is “state somewhere + packets between somewhere.” The tools churn yearly; the two foundations haven’t changed shape in decades — which is exactly why they compound.