--- title: Blog — page 2 of 7 description: Practical guides on monitoring, reliability, and performance — plus product news and lessons from the WatchFor team. canonical: https://watchfor.io/blog/page/2 --- # What we watch for — page 2 Practical guides on monitoring, reliability, and performance — plus product news and lessons from keeping our own systems online. [All](/blog)[Monitoring](/blog/category/monitoring)[Networking](/blog/category/networking)[Performance](/blog/category/performance)[Security](/blog/category/security)[Reliability](/blog/category/reliability)[Email](/blog/category/email)[DevOps](/blog/category/devops)[Engineering](/blog/category/engineering) [All articles](/blog/all) 136 article s · page 2 of 7 [SecuritySep 08, 2026 ## Referrer-Policy Explained: what every outbound link tells strangers Click a link on your site and the browser quietly tells the destination which page you came from — the full URL, including whatever is in it. Sometimes that's a reset token. One header decides how much gets sent.WatchFor Team4 min read](/blog/referrer-policy-explained)[MonitoringAug 23, 2026 ## Six scheduled jobs that will eventually burn you — and the exact check-in for each Backups, certificate renewal, queue workers, ETL, cleanup jobs, billing runs — every team runs them, every team has been burned by one going silent. Recipes with the exact ping patterns: exit codes, run duration, output capture, and the grace period that fits each job.WatchFor Team5 min read](/blog/cron-monitoring-recipes)[NetworkingAug 23, 2026 ## DNS propagation is a myth: what's actually happening when your record 'takes 48 hours' Nothing propagates. DNS changes don't spread across the internet — caches expire, one resolver at a time, on a timer you set yourself. Once you see it that way, migrations stop being scary: here's the TTL playbook, the myths that refuse to die, and the checks that tell you the truth.WatchFor Team4 min read](/blog/dns-propagation-myths)[ReliabilityAug 23, 2026 ## How to read (and write) an SLA report that isn't lying Two companies report 99.95% uptime. One had a flawless month; the other hid a six-hour outage behind measurement tricks. Same number. Here's where SLA reports bend the truth — planned downtime, degraded time, measurement gaps, cherry-picked windows — and what an honest report discloses.WatchFor Team4 min read](/blog/honest-sla-reports)[ReliabilityAug 23, 2026 ## Maintenance windows done right: stop paging yourself for planned work Every team has planned downtime — deploys, migrations, nightly batches. Most handle it by pausing monitors or eating the false alarms. Both are wrong. Here's what a real maintenance window does: silence the noise, keep the data, and keep your SLA honest.WatchFor Team5 min read](/blog/maintenance-windows-done-right)[MonitoringAug 23, 2026 ## The agentic supply chain: monitoring MCP servers you host — and the ones you only consume Every MCP server your agents connect to is a live dependency — like an npm package that can change under you at runtime, with no lockfile and no changelog. Here's the monitoring playbook for both sides of that relationship: the servers you run, and the third-party ones you merely trust.WatchFor Team5 min read](/blog/mcp-agentic-supply-chain)[MonitoringAug 23, 2026 ## Who changed the monitor? Why your monitoring needs an audit trail The postmortem question that has no answer in most monitoring setups: who disabled that alert, when, and what did it look like before? Here's why an audit trail on your monitoring config is as important as one on your production data — for debugging, for teams, and for the SOC 2 auditor.WatchFor Team4 min read](/blog/monitoring-audit-trail)[ReliabilityAug 23, 2026 ## On-call scheduling without a second tool: rotations, escalations and fair shift math Most teams bolt a paging product onto their monitoring and glue them together with webhooks. Here's how full on-call management — rotation calendars, escalation chains, personal paging and compensation-ready shift totals — works when it's built into the monitor itself.WatchFor Team6 min read](/blog/on-call-scheduling-in-watchfor)[MonitoringAug 23, 2026 ## Seven API monitors every SaaS should be running (with the exact assertions) Not a philosophy post — a checklist. Seven concrete API monitors, the exact assertion each one needs, and the failure it catches: the lying health endpoint, the silently expiring token, the queue that backs up on Friday night, and four more.WatchFor Team5 min read](/blog/seven-api-monitors-every-saas-needs)[MonitoringAug 23, 2026 ## The complete status page guide: from bare green dots to a page customers actually trust A status page is the most public thing your infrastructure has. This guide covers every layer of building a good one: branding and themes, components and grouping, automatic vs. manual state, third-party dependencies, maintenance announcements, visibility modes, badges and JSON — and the trust rules that make or break all of it.WatchFor Team7 min read](/blog/status-page-complete-guide)[MonitoringAug 23, 2026 ## Webhook reliability: designing deliveries that survive, and monitoring both ends Webhooks fail silently by design — the sender moves on, the receiver never knows what it missed. Here's the engineering playbook for both sides: retries and idempotency, signatures and ordering, dead letters — and the monitoring that catches a dead webhook pipeline before your data drifts.WatchFor Team4 min read](/blog/webhook-reliability-guide)[MonitoringAug 09, 2026 ## API monitoring beyond status codes: assert on the response, chart what matters A 200 OK only proves your API answered — not that the answer was right. Here's why status-only checks miss real outages, what monitoring the response actually takes, and how WatchFor's new API monitor asserts on JSON, extracts values into charts, and pages you when the body goes wrong.WatchFor Team6 min read](/blog/api-monitoring-beyond-status-codes)[MonitoringJul 30, 2026 ## Heartbeat & Cron Job Monitoring: the Dead Man's Switch Your Backups Deserve Most monitoring checks that something is up. Heartbeat monitoring does the opposite — it alerts when something that should run goes silent. The concept, who needs it, and how to set it up so a dead backup job can't hide for three weeks.WatchFor Team6 min read](/blog/heartbeat-cron-monitoring)[MonitoringJul 30, 2026 ## MCP Server Monitoring: Why "HTTP 200" Means Nothing to an AI Agent MCP servers fail in ways no uptime check can see: the port answers 200 while the handshake breaks, tools vanish, or a tool description is silently rewritten under your agents. Here's how protocol-level MCP monitoring works — and why we built it.WatchFor Team5 min read](/blog/mcp-server-monitoring)[NetworkingJun 26, 2026 ## Advanced SSH: jump hosts, tunnels, and power-user tricks Once you know basic SSH, a whole toolbox opens up: jump hosts to reach private servers, port forwarding, SOCKS proxies, agent forwarding, and a config file that makes it all effortless. The power-user guide.WatchFor Team5 min read](/blog/advanced-ssh-tunnels-jump-hosts)[NetworkingJun 26, 2026 ## FTP vs FTPS vs SFTP: file transfer protocols explained FTP, FTPS, and SFTP all move files — but only two are safe, and they work in completely different ways. Here's how each protocol works, active vs passive mode, the best clients per OS, and which to use.WatchFor Team6 min read](/blog/ftp-ftps-sftp-explained)[ReliabilityJun 26, 2026 ## SLA vs SLO vs SLI: the reliability promise, decoded Three little acronyms quietly run every serious reliability conversation — and almost everyone mixes them up. Here's what SLA, SLO and SLI actually mean, how they fit together, and how to set ones that won't page you at 3am for nothing.WatchFor Team8 min read](/blog/sla-slo-sli)[NetworkingJun 26, 2026 ## SSH explained: secure remote access, keys, and sessions SSH is how you securely log into and control machines anywhere in the world. Here's what it is, how password vs public-key auth really works, how to run remote commands, and the best clients for every OS.WatchFor Team5 min read](/blog/ssh-explained)[ReliabilityJun 25, 2026 ## Alert Fatigue: why your team stopped reading alerts (and how to fix it) When everything alerts, nothing does. Alert fatigue is how real outages slip past tired teams — and it's fixable. Here's why it happens, what it costs, and how to get back to a pager you can trust.WatchFor Team5 min read](/blog/alert-fatigue)[MonitoringJun 24, 2026 ## Status Page Best Practices: turning your worst day into trust A status page is the one place customers look when things go wrong — and most are an afterthought. Here's how to run one that deflects tickets, calms users, and quietly earns trust during an outage.WatchFor Team4 min read](/blog/status-page-best-practices) --- Canonical page: https://watchfor.io/blog/page/2 · Site guide: https://watchfor.io/llms.txt