Insane Node Memory & CPU Consumption

Hi,
I have a node running on quando core celeron with 5GB of RAM, I have a big problem on CPU and RAM usage, the CPU is about 100% all the time:

and the memory the RAM and Swap mem the same:

Looking the CPU graphs I see that the IOWAIT is on 54% average, is that correct ?

What I can do to solve this problem ?

thx

How do you know what’s causing this utilisation? Can you show the processes running and usage by process?

This is the situation 40 minutes after node restart, storagenode use all CPU and RAM:

It’s normal to have high load after a reboot of the node. It starts to calm down because it has to load everything and check files and the database.

The problem is all in this way not jus after a restart (not reboot), take a look the screenshoot posted

no its not normal. my 3tb node never needs more than a few hundred MB RAM and barely any cpu. IOWait is normal though now and then.

Which version are you currently running ?

You do know that everyone nodes aren’t all running on the same spec hardware when I restart my node it uses 100% cpu for awhile. heres my node right after I just simply restart my node.

I’m running 0.34.6 version, I think the last one

The problem is that my CPU is always at 90-100% of usage as for my 5GB of RAM, the graphs attached on my origina post is of 6 hours of range and how you can see is always full CPU and MEM

No I do see it, Which cpu are you running so I can get a better Idea if this is an actual problem or not.

I’m running a quad core Celeron J1900

So heres the main issue ok, Your cpu baseclock is 2.0ghz which older celerons werent know as being a very powerful processor by no means, Lets put this in perspective of how powerful it is, A rpi 3 has a more powerful cpu then this celeron. The TDP of this processor is only 10watts of power.

To me this is going to be a normal issue that your gonna have to get used to with a weak processor. Something that may help a bit If you installed ubuntu on the same drive as the node its going to slow everything down. Installing the OS on a different drive would increase performance a bit because its not going to be sharing the same resources as your node is. Another thing is if your running the OS with a GUI its gonna put alot more overhead vs running without a GUI.

My Node is on a pretty weak cpu as well but I have the OS installed on its own SSD, I am running No GUI, It struggles when start up on one node and if its a fresh reboot I have seen it pegged at 100% cpu for a few hours.

1 Like

The strage things is that the increase of CPU start in February

:

So is this node the same node that you have been running for how long? If I remember correctly there was hardly any activity in Jan. This month alone Im getting almost 100gigs perday again so actvity is pretty high right now.

Yes is the same node since the beginning, the CPU was lower when there was much more traffic that last weeks …so I don’t understand this CPU&MEM usage with lower traffic

So you didnt see high cpu when we were getting lots of data? Or were you not watching it this closely, cause from your graph you didn’t have high cpu/io usage when there was hardly any activity.
What have you changed since dec to now?

Yes lower CPU with higher traffic, I didn’t change any on my node, just regular and automatic node version upgrade

Did you notice this right after you came to 34.6? I havent seen any kinda issues with the update. No memory leaks no constant cpu/io.

This happen before the 0.34.6 release, idk when exactly