We did some digging on Linux memory management recently and came to the conclusion that it's just one big Ponzi Scheme.
=G=
On Fri, Mar 16, 2018 at 1:51 PM, Jason Brockdorf via Xymon <xymon at xymon.com> wrote:
Xymon mailing list Xymon at xymon.com http://lists.xymon.com/mailman/listinfo/xymon
---------- Forwarded message ---------- From: Jason Brockdorf <apocalysque at yahoo.com> To: "'Kris Springer'" <kspringer at innovateteam.com>, <xymon at xymon.com> Cc: Bcc: Date: Fri, 16 Mar 2018 12:51:10 -0500 Subject: RE: [Xymon] real memory status
Here's an answer I found googling...
https://unix.stackexchange.com/questions/147006/real-vs- actual-memory-in-xymon-hobbit
From how I understand it, your system memory will be used by services/apps/etc and that usage is referred to as "actual". The remainder of the memory that's not in use for those purposes will be put to use by the kernel as buffers/cache. Basically in Linux (unless you need to tune this for some specific reason) you really don't need to as worried about "real" memory usage, and the fact that it uses that leftover memory to speed up the system is actually a good thing. As far as xymon works, you should be concerned with the "actual" usage (but not *completely* disregard real usage).
I found a really good article one time that explains it. I'll try to find it and send in reply if I can.
*From:* Xymon [mailto:xymon-bounces at xymon.com] *On Behalf Of *Kris Springer *Sent:* Friday, March 16, 2018 12:18 PM *To:* xymon at xymon.com *Subject:* [Xymon] real memory status
What is the 'Real/Physical' memory status? It always shows above 95% even though my system memory hovers around 30%. I've read that it's based on the results of the 'free' command, but when I manually run that command I only get 'Actual' and 'Swap' values. This 'Real' status doesn't seam 'real' at all. I'm routinely forced to manually ignore that specific value's tolerances via the 'analysis.cfg' file, but that seems counter productive if the 'Real' status actually means something important. Can anyone advise?
Thank you.
Kris Springer
participants (1)
-
solitaryr@gmail.com