Create a Post
cancel
Showing results for 
Search instead for 
Did you mean: 
Contributor

sk168513 - Cpu on first instances high on R80.30 take 210 and higher

Hi!

We did a clean install (upgrade) to R80.30 take 215 on our 23900 appliances (vsx with vsls) three weeks ago.
After two weeks we noticed that we were hit by the sk168513. There is a hotfix for it in take 219, 
but that doesnt seem to work for VSX as mentioned in sk169352...

There is a workaround mentioned in sk168513 though:
fw ctl set int fwmultik_enable_increment_first 1
fw ctl set int fwmultik_enable_round_robin 1

Does anyone know if this workaround also is working and supported on VSX/vsls?
If yes: should it be enabled on the affected virtual-server or on vs0? (fwkern.conf)

Thx

Mattias

 

 

0 Kudos
Reply
6 Replies
Champion
Champion

If I'm reading those two SK's correctly, you need a special hotfix for this issue under VSX that is separate from the hotfix added into R80.30 Jumbo HFA Take 219 for non-VSX gateways.

Gaia 3.10 Immersion Self-paced Video Series
now available at http://www.maxpowerfirewalls.com
0 Kudos
Reply
Contributor

Yes, I agree with you about the hotfix.
Therefore my question if the workaround is working (and is supported) on VSX?

0 Kudos
Reply
Employee++
Employee++

The workaround and (non Jumbo) hotfix are in seperate SKs hence address different problems.

0 Kudos
Reply
Advisor

Applying the below in the affected VS resolves the issue? Anybody tried it and the CPU come down?

fw ctl set int fwmultik_enable_increment_first 1
fw ctl set int fwmultik_enable_round_robin 1

We are facing the same issue after we upgraded the VSX cluster from R77.30 to R80.30 with the Take 219 HF

 

 

0 Kudos
Reply
Employee++
Employee++

For VSX JHF T227 or above is needed.

0 Kudos
Reply
Contributor

We are running VSX and the issue was solved for us after installing take 219
I never tried the workaround.

0 Kudos
Reply