- CheckMates
- :
- Products
- :
- Quantum
- :
- Security Gateways
- :
- [Question] Gateway Optimization
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Mute
- Printer Friendly Page
Are you a member of CheckMates?
×- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
[Question] Gateway Optimization
Hi Checkmates,
Since i dug for more details about CP Security Gateway i found the beauty of CP Gateway. The customize could help improve a customer environment.
CP has lots of capabilities to help optimize the traffic without interrupting operations and taking out security.
im sorry if my query below is covered in training materials or other docs/books, I only check on SK that is accessible to partners and this community.
Regarding SecureXL, these main capabilities help to improve & accelerate traffic and will be offloaded to the secureXL. however i missed one point, how do we know the packet will be offloaded?
- in the existing connection with SecureXL enabled, then the traffic is already accelerated but in the middle of the traffic, there are some malicious packets/payloads. how will the gateway perform?
- is this mechanism could make an impact if we have a red team and make an attack simulation such as syn flooding/pentest?
For CoreXL, if i use the 7K series which has 1x CPUs, 16 physical cores, and 32 virtual cores based on the datasheet. question :
- after enabling CoreXL, is the calculation based on 16 physical cores or 32 virtual cores?
- and then enabling VSX, 1VS using 6 core do the split allocation will allocate by default or need to be set manually? AFAIK, split allocation will allocate automatically with 8 cores. 2 SND & 6 FWK.
Im on trying understand about Dynamic Workflow and MultiQueue, and will back to this thread if i have curiosity 😄
Thanks!
Accepted Solutions
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi @Fabz,
If you have Hyper Threading cores, these are also used. In your case you have 32 cores.
Here I have described some performance tuning points:
- R8x - Performance Tuning Tip - SMT (Hyper Threading)
- R8x - Performance Tuning Tip - Multi Queue
- R8x - Performance Tuning Tip - Dynamic split of CoreXL in R80.40
- R8x - Performance Tuning Tip - SecureXL Fast Accelerator (R80.20 JHF103+)
- R8x - Performance Tuning Tip - Control SecureXL / CoreXL Paths
- R8x - Performance Tuning Tip - SND vs. CoreXL
- R8x - Performance Tuning Tip – HyperFlow
- R8x - Performance Tuning Tip – Lightspeed Appliance (R81.10 + JHF)
You can find more about performance and arichtecture here:
R81.x Architecture and Performance Tuning - Link Collection
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi @Fabz,
If you have Hyper Threading cores, these are also used. In your case you have 32 cores.
Here I have described some performance tuning points:
- R8x - Performance Tuning Tip - SMT (Hyper Threading)
- R8x - Performance Tuning Tip - Multi Queue
- R8x - Performance Tuning Tip - Dynamic split of CoreXL in R80.40
- R8x - Performance Tuning Tip - SecureXL Fast Accelerator (R80.20 JHF103+)
- R8x - Performance Tuning Tip - Control SecureXL / CoreXL Paths
- R8x - Performance Tuning Tip - SND vs. CoreXL
- R8x - Performance Tuning Tip – HyperFlow
- R8x - Performance Tuning Tip – Lightspeed Appliance (R81.10 + JHF)
You can find more about performance and arichtecture here:
R81.x Architecture and Performance Tuning - Link Collection
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
in my case, enabling VSX and split the core based on requirement on each VS. is the HyperThreading still applicable for my case?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Yes Hyper Threading is applicable.
If necessary you should manually assign the CoreXL instances to the HT Core identically.
For example for VS1 with 4 cores:
Core 1 to HT Core 17
Core 2 to HT Core 18
...
That is better for the performance.
