> For those of you managing server infrastructure,
As a developer who has often had to look into problems and performance issues, instead of an infrastructure person, this is basically the bare minimum of what I want to see:
* CPU usage
* RAM breakdown by at least Used/Disk cache/Free
* Disk fullness (preferably in absolute numbers, percents get screwy when total size changes)
* Disk reads/writes
* Network reads/writes
And this is high on the list but not required:
* Number of open TCP connections, possibly broken down by state
* Used/free inodes (for relevant filesystems); we have actually used them up before (thanks npm)
As a developer who has often had to look into problems and performance issues, instead of an infrastructure person, this is basically the bare minimum of what I want to see:
* CPU usage
* RAM breakdown by at least Used/Disk cache/Free
* Disk fullness (preferably in absolute numbers, percents get screwy when total size changes)
* Disk reads/writes
* Network reads/writes
And this is high on the list but not required:
* Number of open TCP connections, possibly broken down by state
* Used/free inodes (for relevant filesystems); we have actually used them up before (thanks npm)