- Products
- Learn
- Local User Groups
- Partners
- More
Quantum Spark Management Unleashed!
Introducing Check Point Quantum Spark 2500:
Smarter Security, Faster Connectivity, and Simpler MSP Management!
Check Point Named Leader
2025 Gartner® Magic Quadrant™ for Hybrid Mesh Firewall
HTTPS Inspection
Help us to understand your needs better
CheckMates Go:
SharePoint CVEs and More!
Hello,
I'm surprized to see that on a couple of brand new 9800 appliances, CoreXL is disable by default (see attached a snapshot of the SysInfo from cpview) ... I wonder whether this could be a consequence / side effect of the new UPPAK mode that is enable by default (maybe some incompatibility?).
We have R81.20 with JHF89 and VSX mode enable.
Any clue?
I suspect i should just enable CoreXL to take advantage of the 40 CPUs of the device ... or would this bring new limitations ?
Thanks in advance for your hint(s)
JB
For VS0 this is OK / recommend per: https://sc1.checkpoint.com/documents/R81/WebAdminGuides/EN/CP_R81_VSX_AdminGuide/Topics-VSXG/CoreXL-...
How many virtual systems are configured and in SmartConsole how many CoreXL instances were provisioned for each?
Dynamic Balancing on VSX does not change CoreXL per VS configuration or otherwise affect VS FWK counts, it only changes the CPU allocation between SND and fwk worker pool.
I believe CoreXL being off with VSX is normal and expected.
For VS0 this is OK / recommend per: https://sc1.checkpoint.com/documents/R81/WebAdminGuides/EN/CP_R81_VSX_AdminGuide/Topics-VSXG/CoreXL-...
How many virtual systems are configured and in SmartConsole how many CoreXL instances were provisioned for each?
Hello Chris,
thanks for your feedback.
I have 2 Vsys configured + 1 Virtual switch.
The 2 Vsys are configured with the default settings for CoreXL, so 1 instance only.
I suppose this is why cpinfo report CoreXL being off for the Vsys others than VS0 (see attached snapshot of SysInfo from VSID 3)
Is there any documentation that could help in sizing the amount of CoreXL instances to configure for a given Vsys?
I guess on the basis on my estimated traffic going to the slow path, I could deduct the number of CoreXL to configure ... but maybe there are more parameters to consider.
Also how could I detect that I am running low on the amount of CoreXL instances ?
Before the users start to shout, I suppose there are some warning signals that my CoreXL instances are high on utilisation.
With VSX off, there are no worries to have with CoreXL instances, from my experience, the number is set / adjusted automatically.
Regards,
JB
@JeromeB it depends of the traffic. I would start with 8 cores for every VS.
But I prefer to use the dynamic balancing feature with R81.20. The system automatically creates SND and fw-worker cores depending on the utilizatíon. See CoreXL Dynamic Balancing part How does it work with VSX?
@Wolfgang , is dynamic balancing really creating and removing FW instances on the fly? I thought his role was "limited" to re-assigning instances between SND and FWK...
In my case, VS1 is now set with 2 CoreXL and I can see the relavant processes (fwk1_0 & fwk1_1) running from the start even no traffic is going through the VS.
[Expert@XXXXX:1]# fw ctl affinity -l -x -flags tn | grep fwk1
---------------------------------------------------------------------
|PID |VSID | CPU |SRC|V|KT |EXC| NAME
---------------------------------------------------------------------
| 29889 | 1 | 0x000000FFFFCFFFFC | V | | | | fwk1_dev_0
| 29896 | 1 | 0x000000FFFFCFFFFC | V | | | | |---fwk1_kissd
| 29911 | 1 | 0x000000FFFFCFFFFC | V | | | | |---fwk1_0
| 29912 | 1 | 0x000000FFFFCFFFFC | V | | | | |---fwk1_1
| 29913 | 1 | 0x000000FFFFCFFFFC | V | | | | |---fwk1_hp
| 29920 | 1 | 0x000000FFFFCFFFFC | V | | | | |---fwk1_service
| 29921 | 1 | 0x000000FFFFCFFFFC | V | | | | |---fwk1_dev_1
Also, the SND are assigned to VS0 apparently and they seem to be pretty static (in numbers). No incrementing / decrementing with the amount of VS and CoreXL configured.
[Expert@XXXXX:1]# fw ctl affinity -l -x -flags tn | grep snd
---------------------------------------------------------------------
|PID |VSID | CPU |SRC|V|KT |EXC| NAME
---------------------------------------------------------------------
| 1724 | 0 | 39 | V | | | E | |---fwk_snd_main
| 1729 | 0 | 39 | V | | | E | |---fwk_snd_dev
| 7978 | 0 | 39 | V | | | E | |---fwk_snd_dev
| 23194 | 0 | 39 | V | | | E | |---fwk_snd_dev
| 23250 | 0 | 39 | V | | | E | |---fwk_snd_dev
| 29891 | 0 | 39 | V | | | E | |---fwk_snd_dev
Dynamic Balancing on VSX does not change CoreXL per VS configuration or otherwise affect VS FWK counts, it only changes the CPU allocation between SND and fwk worker pool.
Like @emmap mentioned, dynamic balancing does not assign worker cores to VS. There's no special assignment of SND cores to a specific VS. You can set your worker cores for every VS, maybay 6 or 8 for both. Without any traffic a default setting is done for the count of SNDs. But if you get more traffic and CPU utilization on the system dynamic balancing will create mor SNDs on the fly if needed and will go back if possible.
Many thanks to you both for the clear explanations
Good to know.
This is normal behavior with VSX, not specific of the 9800. Without VSX CoreXL is enabled by default on a 9800 appliance. If you start cpview in one of your VS you can see CoreXL is enabled. For VS0 there is no need for more then one CPU only management traffic is going through this.
Leaderboard
Epsum factorial non deposit quid pro quo hic escorol.
User | Count |
---|---|
11 | |
7 | |
7 | |
6 | |
6 | |
5 | |
5 | |
5 | |
5 | |
4 |
Wed 10 Sep 2025 @ 11:00 AM (CEST)
Effortless Web Application & API Security with AI-Powered WAF, an intro to CloudGuard WAFWed 10 Sep 2025 @ 11:00 AM (EDT)
Quantum Spark Management Unleashed: Hands-On TechTalk for MSPs Managing SMB NetworksFri 12 Sep 2025 @ 10:00 AM (CEST)
CheckMates Live Netherlands - Sessie 38: Harmony Email & CollaborationWed 10 Sep 2025 @ 11:00 AM (EDT)
Quantum Spark Management Unleashed: Hands-On TechTalk for MSPs Managing SMB NetworksFri 12 Sep 2025 @ 10:00 AM (CEST)
CheckMates Live Netherlands - Sessie 38: Harmony Email & CollaborationAbout CheckMates
Learn Check Point
Advanced Learning
YOU DESERVE THE BEST SECURITY