On Fri, Mar 01, 2019 at 07:34:50PM +0100, Pascal Hambourg wrote: > Le 01/03/2019 à 18:56, Reco a écrit : > > > > First, there's huge amount of unused (not to be confused with "free") > > memory on your host. And no, it's not a filesystem's cache (600M), it's > > really "nothing there"-unused memory which amounts 1880M. > > This is the amount reported in /proc/meminfo as MemFree - not MemUnused. > Can you please explain your definitions of "free" and "unused" memory ?
Documentation/filesystems/proc.txt.gz What they call "free memory" usually is called MemAvailable in /proc/meminfo. "MemFree", which was the sum of "LowFree" and "HighFree" back in the day (and still is for some architectures) - is the amount of memory which isn't used for anything, hence the "unused". > > Taking it all into the account, I propose the following scenario - not a > > long time ago there was a process (or a couple of those). Was it the > > browser, or something written in Java (or Python), or some kind of > > Virtual Machine - is not relevant, but it did consume RAM. But it got > > terminated, and what we're seeing here is the aftermath. > > Another scenario : at some point the pagecache (file cache) grew much > bigger due to massive file read/write operation, then the files were > deleted or the filesystem was unmounted and the file cache pages were > freed. I agree that's another possibility. Reco