TrueNAS-SCALE-22.02-RC.1-2 high iftop cpu usage.

[o1crimson1o]

7132 root 20 0 233316 5596 3216 R 101.0 0.0 2:03.16 iftop
7126 root 20 0 233316 5620 5264 R 100.0 0.0 2:03.21 iftop
7128 root 20 0 233316 5680 5324 R 100.0 0.0 2:03.18 iftop
7130 root 20 0 233316 5712 5360 R 100.0 0.0 2:03.21 iftop
7133 root 20 0 233316 5544 3168 R 100.0 0.0 2:03.14 iftop

Which keeps the load avg on the machine up to 5-8, I'm more concerned about loosing so many cpu cycles to simply monitor b/w.

18:00:26 up 5 min, 1 user, load average: 5.69, 4.52, 2.13

Version: TrueNAS-SCALE-22.02-RC.1-2

I don't have this issue on Truenas.

I'm not sure what else to

 

[o1crimson1o]

I can't edit but to follow up, this is a fresh install, on a 4 port LACP setup.

 

[morganL]

TrueNAS SCALE has a different linux implementation.
Performance testing is just starting with RC2.... so you may see inefficiencies like this.

 

[Neo1711]

Any chance to get rid of this until officially fixed?
My box is running at 100% CPU (a couple of NICs and VLANs) and slows down everything...

 

[NugentS]

Its RC software. Its gonna have bugs.
If you think its a bug then raise a JIRA ticket and they will get on it.

 

[Neo1711]

Its RC software. Its gonna have bugs.
If you think its a bug then raise a JIRA ticket and they will get on it.

Done => https://jira.ixsystems.com/browse/NAS-113781

 

[t3chn0g33k]

Having the same issue with SNMP enabled on TrueNAS-SCALE-22.02-MASTER-20211228-012923.

What is the proper steps. Create another bug or is there a way to add to the one listed above (https://jira.ixsystems.com/browse/NAS-113781)?

 

[Neo1711]

Having the same issue with SNMP enabled on TrueNAS-SCALE-22.02-MASTER-20211228-012923.

What is the proper steps. Create another bug or is there a way to add to the one listed above (https://jira.ixsystems.com/browse/NAS-113781)?

Hi,
Ticket is open and being processed.
Within the discussion we have identified that this only happens when SNMP is enabled.
IMHO they are working on that...

 

[rickyrickyatx]

I temporarily 'solved' this issue by disabling iftop data being integrated into net-snmp for anyone else wondering how to stop if top from eating up a full CPU's worth of availability (per interface it seems like).

 

[Neo1711]

Bug has been fixed in SCALE 22.02.0
https://www.truenas.com/docs/releasenotes/scale/22.02.0/