fw_full process high CPU usage standy member

[Amodi]

Hi, we have a HA cluster of two R80.30 using take: HOTFIX_R80_30_JUMBO_HF_MAIN Take: 236

After switching the current active member into standby, it's $FWDIR/log/fwd.elg is flooded with the following lines:

CCplogUtils::CreateStringDefLog error
CCplogBoCommClientCommon::SendStringDefinition : error: fail to create string def log
CCplogBoCommClientCommon::SendBinObj : error: fail to send string definition

and the fw_full process is topping the CPU.

After a restart or cpstop/cpstart the flooding message will stop, and so will the CPU usage.

Does anybody had seen this happening, and eventually has a solution beside the restart?

Thanks,

--Sanyi

[PhoneBoy]

Best to engage with the TAC to assist in troubleshooting this.

[Amodi]

Thanks PhoneBoy, I'll take your suggestion, and engage with the TAC.

--Sanyi

[Henrik_Noerr1]

We see this as well - could you please share the findings from TAC? r80.30 take 215

Regards

Henrik

[Amodi]

Unfortunatelly I had no time to engage with TAC, but plan to make some, in order to clear this issue.

If I have any update, I will post it here.

[Lukas_Sosnovec]

Hi,

I have the same problem on gaia embeded R80.20.30 - Build 285
I am working with TAC on it but no results yet...

Did you make any progress?

[PhoneBoy]

I presume the issues for this are different in Embedded Gaia.
Please create a post in the Quantum Spark section with the relevant details. 

[Amodi]

No updates, and hope that this issue does not exist in R80.40, as I plan to update our systems.

[Amodi]

Just updated the gateways to:  Check_Point_R80_30_JUMBO_HF_Bundle_T237_sk153152_Security_Gateway_and_Standalone_2_6_18_FULL.tgz

and I didn't see the high CPU usage on the standby member anymore.

[pankajjain1]

Hi All,

I have a same problem in r80.40 FWD CPU process is showing high and in top command I am seeing the fw_full process is taking 100% CPU which is related to fwd process.

Gaia r80.40

Take_156