Load averages
-
hm - i have a similar question:
What is the correct way to walk if one likes to contribute a feature ? -
You can see detail for load averages and such on Diagnostics > System Activity.
If you want to contribute a change, it's easy to sign up on github, fork the pfsense repo, make your edits, and then submit a pull request. Github has some good documentation that covers the whole process.
-
You can see detail for load averages and such on Diagnostics > System Activity.
Yes I know I can see the load averages on the "top" page, but IMO they are much more useful than (in example) the CPU activity meter. Taking a quick look at those load averages will tell you a lot more on how the system is working than a CPU meter that changes every 10 seconds. That's the reason why I think they should be on the System Information widget. Now the question is: If I write the code and make the pull request, will it get approved? :)
-
If it's a simple change and easy to review, sure. We take contributions all the time. Something like that could be added into the existing system information widget just below the cpu meter perhaps.
-
If it's a simple change and easy to review, sure. We take contributions all the time. Something like that could be added into the existing system information widget just below the cpu meter perhaps.
Great! Maybe above the CPU meter would be better so it won't break up the look of those four graphs?
I'll get into it as soon as I can, any idea on when the final 2.1 is being released? -
We want at least a beta of some sort out in time for World IPv6 launch, so maybe a month or two.
-
If it's a simple change and easy to review, sure. We take contributions all the time. Something like that could be added into the existing system information widget just below the cpu meter perhaps.
Great! Maybe above the CPU meter would be better so it won't break up the look of those four graphs?
Why not continue the same line that has the CPU load? That would mean one line shows current CPU utilization, and averages, which means the whole thing doesn't take up any extra space.
-
Oh sorry too late, I already commited it with a new line :(
It doesn't use much extra space anyway and looks cleaner this way, with the proper title at the left.Why not continue the same line that has the CPU load? That would mean one line shows current CPU utilization, and averages, which means the whole thing doesn't take up any extra space.
-
approved the request, looks good to me (though my test vm is always load 0.00 :-)
-
Nice, that was fast.
Try reloading quickly a few times with the browser, that worked for me ;)approved the request, looks good to me (though my test vm is always load 0.00 :-)
-
Hi feadin - fetched the latest snap Yesterday wt your load display
maybe you know that im playing with the temperature widget ?
I tried if the bar would work with your meter and it does :)Do you want me to adapt it to a bar ? would look like this:
-
The issue here would be… exactly what is "100%"?
The load average values per se are a very common sight for a system administrator, very easy to see and understand in one quick look how things are going. But they are not so easy to translate to a fixed percentage value (at least IMO).
I don't know if you're familiar with load average values, so just in case here is a nice and simple explanation: http://blog.scoutapp.com/articles/2009/07/31/understanding-load-averages@ThorstenK:
Hi feadin - fetched the latest snap Yesterday wt your load display
maybe you know that im playing with the temperature widget ?
I tried if the bar would work with your meter and it does :)Do you want me to adapt it to a bar ? would look like this:
-
@ThorstenK:
Hi feadin - fetched the latest snap Yesterday wt your load display
maybe you know that im playing with the temperature widget ?
I tried if the bar would work with your meter and it does :)Do you want me to adapt it to a bar ? would look like this:
There's another problem: you collapse the three values into one, if I understand correctly, and it's often useful in tracking down a problem to see if one's dealing with spikes or whatnot.
If a graph, it would be better the kind like used for the traffic graphs, at which point one could probably get rid of the CPU usage bar, too.
-
Maybe that will sound a bit strange, but im not used to such high load situations -ok, since im using vm player for ages now, im used to just setup multiple vms. (lets dont hang on that demo, even 3 bars wt values are no problem). Just had the code in my fingers and it was an idea i wanted to share. hm - but i like the idea of playing with the traffic graphs. will see :)
