performance problems and disconnections with R81.20 JHF 53

[Flo]

Hello everyone,

 I am now experiencing massive performance problems and disconnections, especially to SAP, on the 2nd cluster after upgrading to R81.20 JHF 53.

 Is anyone else experiencing this?

 TAC case for both cases is open.

[Lesley]

Can you make it more specific? From what version and take you had before update.

See any high load(cpview) or drops in logs

[Flo]

Hey @Lesley 

on the first environment I have updatet from R81.00 to R81.20 JHF 53

The second environment I have updatet from R81.20 JHF 41 to R81.20 JHF 53

So it seems to me, that the JHF 53 is the problem maker...

In /var/log messages I see some spikes, and in the spike datas I found entries like native_queued_spin_lock_slowpath

But I do not know, if that is resplonsible for that problems

[Lesley]

No take installed on R81 before? That was a big step to upgrade 🙂

Maybe start with a HCP (health check report). Maybe we can find something in there. Run it on both (active) gateways. 

https://support.checkpoint.com/results/sk/sk171436

[Flo]

HCP is all fine.

[Chris_Atkinson]

What is the gateway model in each case and which sort of disconnections are you referring to, is the issue always present or only when installing policy?

[Flo]

Thank you for your support.

I think this is going in the wrong direction... 

It was just a general question if there are more people in the community who have the same problems, as I'm on my 2nd upgrade now, with similar errors.

If that's not the case, then maybe my first guess wasn't correct and it just happens to be the same Jumbo with a similar error pattern...

[Flo]

we can't really grasp the issue yet, as nothing unusual can be seen in the log or in the zdebug.

The issues are always there, not just during the policy install