Memory Spike Reason Unknown

stephenw10

Nothing there looks too huge when that was taken.

FreeBSD commands should work, it's just FreeBSD at the command line. What are you trying to run?

Steve

RJP1267

It spiked back up to 33% and am looking for the culprit, I think it may be squid related.

stephenw10

Squid can use a lot of RAM especially if you have non-default local cache settings. How much RAM do you have on that box?

RJP1267

@stephenw10 4gb...what do you suggest to ensure not over using resources when not needed? Thank you.

Gertjan

@rjp1267

Try this one :

pkg install htop

Now you have htop at your dispossal ;)

And a question : when I sort on memory usage :
Who is this :

It's also rather new here :

( I know, Google shuld be my friend .... )

stephenw10

See: https://redmine.pfsense.org/issues/12095

SteveITS

Interesting, on a 3100 it shows as:

Mem: 31M Active, 1156M Inact, 199M Wired, 83M Buf, 594M Free

PID USERNAME PRI NICE SIZE RES STATE C TIME WCPU COMMAND
13149 root 20 0 960M 950M nanslp 1 32:47 0.06% /usr/local/sbin/pcscd{pcscd}
13149 root 20 0 960M 950M select 1 1:19 0.00% /usr/local/sbin/pcscd{pcscd}
13149 root 52 0 960M 950M piperd 1 0:00 0.00% /usr/local/sbin/pcscd{pcscd}

Yet the dashboard shows "Memory usage 12% of 2017 MiB" (242 MB).

On a PC with 2.5.2 it shows as:
Mem: 87M Active, 2768M Inact, 972M Wired, 592M Buf, 3296M Free
Swap: 3852M Total, 3852M Free

PID USERNAME PRI NICE SIZE RES STATE C TIME WCPU COMMAND
7457 root 20 0 1946M 1928M nanslp 3 21:30 0.08% /usr/local/sbin/pcscd{pcscd}
7457 root 20 0 1946M 1928M select 2 0:30 0.00% /usr/local/sbin/pcscd{pcscd}
7457 root 52 0 1946M 1928M piperd 1 0:00 0.00% /usr/local/sbin/pcscd{pcscd}

and the dashboard memory usage "15% of 7314 MiB" (1097.1).

The dashboard must not be counting it?

stephenw10

Yeah, in most cases it doesn't actually seem to cause a problem. But it shouldn't behave like that and it will not be enabled by default in the next release.
https://redmine.pfsense.org/issues/11933

Steve

RJP1267

@gertjan Now that htop is the bomb, great call. Now if I can find info packed dashboard alternatives I will be on my way.

RJP1267

Using htop I was able to see that 5+% of the 33% mem being used was in use by squid...Is that normal in a basic out of the box config (1 wan IF and 1 Lan IF <12 internal ip addresses)?

stephenw10

Too many variables to say what's 'normal' there really. But unused RAM is pointless so something using 5% of 4GB would not concern me.

Steve

RJP1267

@stephenw10 Yeah makes sense, what caught my attention was the % used in the dashboard, I used to run at 18% then saw 33 made me wonder. I can really do with a new dashboard, I saw some discussions on other boards about them but they seem time consuming to build but they looked sweet and you can really get creative with what is displayed.