- This topic has 6 replies, 4 voices, and was last updated Feb 20-12:30 am by mikey777.
-
Posts
-
mikey777
In wanting to measure the RAM activity of one of my antix setups, I noticed that different methods gave different values.
This presented a conundrum: are they measuring RAM in different ways and, if they are, which method gives the truest actual value?
This in itself is difficult, as the actual measurement in itself can affect the thing you are measuring (familiar to anyone say in
studying animal behaviour), e.g. just opening the terminal before typing anything may use up RAM activity, though this seems is
generally small (~ 4-5MB for xterm or uxterm; up to 12MB for LXTerminal) .Here are the results of measuring idle RAM activity using 3 different methods on a 32bit antix19.1core+LXDE full desktop setup on a single-core Intel Atom netbook (kernel 4.4.212). Following the ‘launch’ of a particular method using terminal, the system was left to stabilise for 5 minutes before recording the data, to give consistency between comparisons. Terminal output data for the top and inxi -m methods are given below in ‘MiB’ or ‘GiB’, and I assume the same units of measurement are used for < 5 minutes, em>htop, which abbreviates units of measurements to ‘M’ and ‘G’:
top ……………… 96 used / 2006.7 total
htop …………….. 103 used / 1.95 total
watch -n 2 inxi -m … 115 used / 1.96 total
ps_mem.py ………… 147 used / (total not given)So guys, which one represents the most accurate measurement of RAM, given there is such a large range from 96 to 147 MiB ? A bit puzzling to say the least …
- This topic was modified 3 years, 2 months ago by mikey777.
- This topic was modified 3 years, 2 months ago by mikey777.
- This topic was modified 3 years, 2 months ago by mikey777.
▪ 32-bit antix19.4-core+LXDE installed on :
- (2011) Samsung NP-N145 Plus (JP04UK) – single-core CPU Intel Atom N455@1.66GHz, 2GB RAM, integrated graphics.
▪ 64-bit antix21-base+LXDE installed on:
- (2008) Asus X71Q (7SC002) – dual CPU Intel T3200@2.0GHz, 4GB RAM. Graphics: Intel Mobile 4 Series, integrated graphics
- (2007) Packard Bell Easynote MX37 (ALP-Ajax C3) – dual CPU Intel T2310@1.46GHz, 2GB RAM. Graphics: Silicon Integrated Systems.
anticapitalista
You forgot
sudo ps_mem.pyI guess they measure RAM usage differently, which is why the numbers vary.
Philosophers have interpreted the world in many ways; the point is to change it.
antiX with runit - leaner and meaner.
mikey777
You forgot
sudo ps_mem.py
I guess they measure RAM usage differently, which is why the numbers vary.Thankyou anticapitalista for that one, which I’ve added to the above.
However when I used …
sudo watch -n 2 ps_mem.py
I couldn’t get the terminal to scroll down to the end to observe the continuous monitoring of the total RAM usage value, when using Shift + PgDn keys. So, instead I’ve just given a rough average of a few ‘static
values’. Using the vertical scrolling-bar on the right side of the terminal failed to work too. I looked
in terminal Preferences but it’s already preset to show a 1000 lines. I’m not sure how to fix this …▪ 32-bit antix19.4-core+LXDE installed on :
- (2011) Samsung NP-N145 Plus (JP04UK) – single-core CPU Intel Atom N455@1.66GHz, 2GB RAM, integrated graphics.
▪ 64-bit antix21-base+LXDE installed on:
- (2008) Asus X71Q (7SC002) – dual CPU Intel T3200@2.0GHz, 4GB RAM. Graphics: Intel Mobile 4 Series, integrated graphics
- (2007) Packard Bell Easynote MX37 (ALP-Ajax C3) – dual CPU Intel T2310@1.46GHz, 2GB RAM. Graphics: Silicon Integrated Systems.seaken64
I usually use “free”. And I’m not usually too worried about accuracy, but in the relative comparison from one setup to another. If I stay with “free” on all the comparisons I get a good idea of the relative effect of my setups. And I always take readings under the same circumstances between each session – i.e. idle at bootup, or with SeaMonkey only loaded with the same web page loaded on a single tab, etc.
Seaken64
- This reply was modified 3 years, 2 months ago by seaken64.
Brian Masinick
If you use a tool that shows the amount of memory used for each process you can get a good idea of which process is using the most memory.
Also there is virtual memory, shared memory and physical memory. If you are familiar with each of them a tool like top or htop may be helpful.
Hope our comments are helpful. If not please let us know what else would be useful to you.
--
Brian Masinickmikey777
If you use a tool that shows the amount of memory used for each process you can get a good idea of which process is using the most memory.
Yes, ps_mem.py seems to be ideal for this. However, do you know though how the continuous monitoring issue, mentioned in #32784, can be fixed?
Also there is virtual memory, shared memory and physical memory. If you are familiar with each of them a tool like top or htop may be helpful.
I’m not very familiar with these. All I have simply been using, for comparing antiX editions on different machines, is the Mem readout in htop – see attached screenshot below. What does this correspond to ?
Hope our comments are helpful. If not please let us know what else would be useful to you.
Many thanks for your offer of help – this is really appreciated. Even though I’ve been using Linux for 5 years, I still feel like a newbie. There seems to be so much more to know, it kinda seems never-ending … But then again, I suppose that’s the fascination with it, there’s always something more to discover.
- This reply was modified 3 years, 2 months ago by mikey777.
Attachments:
▪ 32-bit antix19.4-core+LXDE installed on :
- (2011) Samsung NP-N145 Plus (JP04UK) – single-core CPU Intel Atom N455@1.66GHz, 2GB RAM, integrated graphics.
▪ 64-bit antix21-base+LXDE installed on:
- (2008) Asus X71Q (7SC002) – dual CPU Intel T3200@2.0GHz, 4GB RAM. Graphics: Intel Mobile 4 Series, integrated graphics
- (2007) Packard Bell Easynote MX37 (ALP-Ajax C3) – dual CPU Intel T2310@1.46GHz, 2GB RAM. Graphics: Silicon Integrated Systems.
- You must be logged in to reply to this topic.
