Just did an upgrade from 4.11 to 5.0rc5 on a RB750G and noticed that the CPU is 98-100% all the time. There is no traffic passing while this is the case. No VPN, no proxy, simple ruleset, two queues that are idle.

Any ideas ?

I noticed there is a firmware update available from 2.25 to 2.28


Cheers.

Yes, I would try to upgrade the firmware and see if it makes any difference. If it doesn't, what does the top command reveal? If all else fails, send a supout to support.

maybe you have proxy or dude package installed?

Noticed this release has similar effect on RB600 - CPU usage has jumped from approx 30% average to 75% despite no other changes

rc5 has the profiler tool. go and see what exactly consumes the CPU

A process called 'idle' seems to be using 50-75% of resources. What is that?

A process called 'idle' seems to be using 50-75% of resources. What is that?

idle means that such amount of CPU is available for use (not used)

Seems like the Dude is seeing available resources as CPU usage perhaps?

1. http://www.answers.com/topic/idle
2. which OID are you using for CPU usage? I don't see such problems with my Dude

Seems counterintuitive to show available resources as a resource being consumed then, doesn't it?

I've no idea which OID other than the standard probe that the Dude makes on a Mikrotik device. How can I find out?

Actually comparing CPU in Winbox and the Dude shows the same pattern - so its not the Dude...

Ok I will test.

The Profiler shows total of 100%. For example 30% used, 70% idle. Basically this means that you don't have to calculate how much free resources you have left if you are using 67.33% of the CPU time

That is standard way to display resources, nothing that we invented.

Thanks Normis

It looks like CPU figure is showing total of profile but maybe should be subtracting the idle component.

interesting screenshot:

Can you show me a screenshot with the Dude showing a different CPU usage than Winbox profiler? I can't find any differences (other than slow update speed in the Dude) but maybe I'm looking in the wrong place.

I looked at the profiler and it showed SNMP and management were eating up all my CPU. I removed the SNMP settings and the CPU usage dropped to 1%. I then re-enabled SNMP and it remained at 1%.

Thanks for all the help and suggestions.

interesting screenshot:

well, you will use a rubbish OS!

I can confirm that rc5 is eating up 100% of the CPU ... it doesn't always happen, but when it does, it stays like that for hours.

This is a new rc5 installation on dual core x86 (2x1.8Ghz) with no traffic at all.
The idle process shows 50% utilization, but that's only because of the dual core ... one of the cores is at 100% and it is not illusion or false 100% representation. The CPU does heat up and the cooling fan goes into a higher RPM.

Geza

as you can see, apart from idle, you have management and unclassified. send supout.rif file to support and we will see what is that 'unclassified' process.

My have a same problem. Sometimes CPU usage 100% and it continues about 2 minutes. When I disable SNMP, CPU usage decreases to 20%. It happens every 20minutes. What is that management? I have 2pcs RB1000 V5.0rc4. All RBs have same problem.

I found the problem cause only when I try to manage with Dude server. Other SNMP monitoring programs work fine with my RB1000. MT Support advises me to upgrade ROS version. But still I coudn`t do that. Because it`s not easy upgrade router any time. Our service is 24/7. Anyway now it works fine.

I have v5.5 and my "management" process put my RB800 CPU to 100% every few minutes. When I disable SNMP then menagement is from 0-5% and CPU never goes up to 100%. I monitor my router with Dude.

I have the same problem on my RB133 i posted topic here:

http://forum.mikrotik.com/viewtopic.php?f=1&t=52915

I attached screenshots from cacti.

I have same problem with v5.7

I have same problem with v5.7

which same problem?

monitoring two routers with dude causes 100 cpu usage . the profiler shows it is management . after disabling snmp problem solves . again enabling snmp made cpu usage 100% .

Hi Guys,

I'm having the same issue where SNMP and WebProxy is disabled. No DUDE agent installed or monitoring just a RB doing NAT and IPSec, very simple setup.

See the screen shot below and let me know if there's anything we can do to fix this high CPU usage on the management profile.

Thanks,
Jakub

make a supout.rif file and send to support. make the file at the moment when management is so high.

I'll be sending the support file in a few minutes. Here's what I found, once the IPSec tunnel get's established to another RB750G the CPU goes through the roof. There's only very little traffic (about 350kbps). I tried to decrease the encryption to the lowest possible setting and still an issue. We did a test last night when there were no users on on the network, CPU still at 100% after IPSec tunnel get's established. The master side RB750G has 5 IPSec tunnels and CPU is on avg 30-60%. We hare ordered a RB493G as a replacement, but will wondering why the CPU would go high right after the RB493 establishes the runnel.

Hi Gents just a quick update the RB493 high CPU was killing the router soo much I can't even make a support file, it times out. No file is created. Sorry I wasn't able to provide further diagnostics, but IPSec was an issue with the management in profiler. We replaced the RB493 with RB493G (same config) and CPU is below 8% on average.

any one can confirm that this problem solved in v5.9 ?

I guess this is the right thread to share my findings.

I replaced old RB411 with brand new RB411UAHR today, because it was restarting quite often (by watchdog timer) and it was always preceded by huge cpu usage by management process (and I also needed the second radio). As RB411 has only slow CPU, it was at 100%, with management getting as much as it could and other processes used almost nothing. With faster RB411UAHR it was a little different, management was taking ~70%, snmp ~20% and 10% was idle. The relatively big share for snmp made me investigate it.

And sure, in my case it's definitely SNMP-related. When it was happening and I either disabled SNMP on RB or blocked incoming SNMP packets by firewall, it immediately stopped and router went back to >95% idle. Another thing I quickly found out, was ~500 incoming SNMP packets every second (usually it's 10-20 every few seconds). SNMP logging on RB revealed that queries produce some error and Dude is trying again and again in infinite loop. Unfortunately I didn't write the error down before I "fixed" it, but it was something like invalid or missing value or something not found. What I know exactly, it was happening in DHCP leases table (.1.3.6.1.2.1.9999.1.1.6.4.1). When I tried to snmpwalk RB using different tool, it also got stuck in same infinite loop at OID 1.3.6.1.2.1.9999.1.1.6.4.1.4.192.168.82.50. Manual snmpwalk in Dude did not get stuck, but under dhcpv4ServerClientEntry the only item shown was dhcpv4ServerClientLeaseType and all others usually present were missing.

The "fix" in my case was to remove all dynamic DHCP leases and disable and then enable all static ones. After that, it doesn't happen any more. Well, so far... but I'm pretty sure it will come back when least needed.

ROS 5.11 (and few older 5.x before that), The Dude 4.0beta3. Not sure whose fault it is.

I have had this same issue since September of last year when I updated from 4.17 to whatever version 5 there was at the time. I was met with similar condescending rude responses from mikrotik support. This seems the norm from them. I'm not sure if this is because they are not native english speakers, but all through this forum it's full of condescending responses just like what he said about how they didnt invent how idle processes are shown, even though it was not at all relevant to what was said. Normis misunderstood the question, yet still had no problem being rude and condescending about it.

Anyways, When I inquired about this issue, they told me it was a graphical error and since the router still responded fine to pings, and did not show poor performance, not to worry about it. I sent support files as well.

I just turned off SNMP and the cpu went from 100% to 30% for the first time on the graphs since September.

tl;dr;
I turned off SNMP and all the "management" cpu usage went away.

Would have been nice to know this back when the issue started last year.

Please try to install RouterOS v5.14, if you still see the same problem contact support with attached supout file.

Please try to install RouterOS v5.14, if you still see the same problem contact support with attached supout file.

We are also seeing this on v5 but not on v4. Can you confirm if Mikrotik was able to replicate this issue and if v5.14 included a fix? No mention in the changelog so any information greatly appreciated.

i have same problem..look at a screenshoot..

some times a load high use bt management profile, but now is use by unclassified profile.

Update your ROS to 5.24 if it happens again generate supout file and send it to support. Even you can try 5.25rc1.