Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


How to interpret the new beancounters scheme?
New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

How to interpret the new beancounters scheme?

yomeroyomero Member
edited September 2011 in General

At the end, we know how to read the old OpenVZ scheme, but with the new memory management it seems different.

My current research says:

  • privvmpages: The real memory used
  • physpages: The memory+buffers/cache used, but probably physpages + shmpages is more accurate to memory+buffers/cache

Seems like that, but I am not sure. Also, I don't know what means oomguarpages or if some of these counters is for the swap.

The OpenVZ wiki seems outdated (or I haven't found a better article about UBC)

Comments

  • I thought new OpenVZ shows memory/buffers/swap correctly now

    # free
                 total       used       free     shared    buffers     cached
    Mem:        131072      79536      51536          0          0      64444
    -/+ buffers/cache:      15092     115980
    Swap:       131072       7224     123848

    Or I'm completely wrong? Honestly, I haven't really researched it much xD

  • Well, it seems, but I can't be sure because OpenVZ still doesn't load modules like other virtualization environments. And yeah, seems like is properly counted now.

    But I posted this because must exist an interpretation for that counters, I think.

    And a side question is:

    Under this new model, how can we detect overselling?

  • yomero said: Under this new model, how can we detect overselling?

    How do you define "overselling" for an OpenVZ provider? :) If you mean your allocated memory getting swapped, you can still check whether the held value of physpages is the same as oomguarpages, to determine how many of your committed pages are still in physical memory.

    After all vswap is just a "virtual swap", which is created artificially to be slow to access. It's possible your swapped pages are still in physical memory, and it is also possible your committed pages have actually been swapped out on the physical machine.

Sign In or Register to comment.