Hardware monitoring watches the machine itself, temperature, disk health, power, independent of whatever your application or OS reports about itself.
What is monitored below the OS
- Disk health (SMART attributes): reallocated sectors and other early-warning signs a drive is degrading, often visible well before an actual failure.
- Temperature and fan speed: sustained high temperature is one of the most common precursors to hardware failure.
- Power supply status: on servers with redundant PSUs, a single failed unit that has not yet caused an outage.
- Network link status: a downed physical network interface, distinct from a software-level connectivity problem.
This layer runs independent of your operating system, through the same out-of-band management chip IPMI uses, so it keeps reporting even if the OS itself is unresponsive or the server will not boot.
What it does not cover
Hardware monitoring has no visibility into your application, your database, or your own server's CPU/memory usage from the inside, that is a separate layer you set up yourself (a monitoring agent, uptime checks against your own endpoints). The two layers are complementary, not redundant, hardware monitoring catches problems your application-level monitoring cannot see coming, and vice versa.
How you are notified
A hardware alert on your dedicated server opens a proactive notice from our side, often before you would notice anything yourself. A degrading disk flagged this way gets scheduled for replacement before it fails outright, which is the entire point of catching it at the SMART-warning stage instead of at the actual-failure stage.
The value of hardware monitoring is entirely in catching things before they become outages. A disk that fails with no warning is rare, one that gave weeks of warning nobody was watching for is common.
Can't find something?
Open a ticket and a real engineer will reply, usually inside two minutes.
Open a ticket