High CPU usage on 'rate' service
I noticed recently that the CPU usage on my pfSense system was looking quite high - running at something between 50 to 80 percent constantly. Running top, I saw that the 'rate' service, which I believe is what generates system graphs, was using up a significant chunk of this. As an experiment, I killed the PID for the rate service and immediately the CPU dropped back down to something more reasonable.
I have both Googled this problem as well as consulted other posts on this forum, but most seem to suggest that it was a problem with the old 1.2.3 release. I'm presently running 2.1.5 and it seems the issue is still present. Is there any updated information on this problem which I'm unaware of? Is this an existing bug which still hasn't been addressed or is there a workaround or fix that can be applied apart from just killing the service?
Regards to all
Sorry I can't shed any light on the graphing CPU issue directly, but is there any reason not to upgrade to 2.2.2?
I'm thinking that debugging an issue from the latest stable is liable to get more resources than 2.1.5 will.
Other than some possible funkiness with particular hardware (hasn't bit me yet) and one required physical reboot (that did get me) the changeover to 2.2.2 has been very smooth for me. I've done some 20+ systems remotely and have only hit one issue with an old piece of thin-client hardware that I'm fairly certain I did to myself in an oddball install of 2.1.5. Overall I'm very happy with 2.2.2, I think it's worth a try.
Just my $.02, YMMV
Thanks for the reply. Since stopping the service I've not had a reoccurence of the CPU problem. I will see about upgrading shortly, though I have to be careful as this is a production system. I guess I'll have to just see whether the upgrade solves the issue eventually.