Trunk hash mode detail

[mr_vaughn]

We have many clients with HP Procurce 25xx, 54xx & 29xx switches. We are starting to use the Netgeat M4300-24X Switches and uplink all the HPs via trunks/LAG's to it as Core switch.

I know the HPe side has 3 types but it does not go into detail as to how each works. trunk-load-balance <L2-based | L3-based | [L4-based>

As per https://techhub.hpe.com/eginfolib/networking/docs/switches/RA/15-18/5998-8161_ra_2620_mcg/content/ch04s11.html

For the netgear on LAGs we have the following options

• Src MAC, VLAN, EType, incoming port- Source MAC, VLAN, EtherType, and incoming port associated with the packet.
• Dest MAC, VLAN, EType, incoming port- Destination MAC, VLAN, EtherType, and incoming port associated with the packet.
• Src/Dest MAC,VLAN,EType,incoming port- Source/Destination MAC, VLAN, EtherType, and incoming port associated with the packet.
• Src IP and Src TCP/UDP Port Fields- Source IP and Source TCP/UDP fields of the packet.
• Dest IP and Dest TCP/UDP Port Fields- Destination IP and Destination TCP/UDP Port fields of the packet.
• Src/Dest IP and TCP/UDP Port Fields- Source/Destination IP and source/destination TCP/UDP Port fields of the packet.
• Enhanced hashing mode- The purpose of hashing based load balancing is to split micro-flows in a way that the traffic load is balanced across multiple output links.

Question is what matches the HP Procurve default of L3 based from the NetGear 7 options?

[parnassus · Accepted solution]

Hi!

---

@mr_vaughn wrote: Question is what matches the HP Procurve default of L3 based from the NetGear 7 options?

---

Why to care about it?

The hashing mode is acting (and has effect) only on egressing traffic from a LAG (Port Trunk in HP "E-Series" jargon)...so your HP switches would adopt the hashing algorithms you think are better suited for the type of egressing traffic they have with destination your Core Switch and vice-versa your Netgear Core will do the same for its egressing traffic with destinations its downlinked peer switches...and will do that freely using its hashing algorithms (Note: the 6th on Netgear list should be equivalent to the Layer 4 = L4 on HP).

Hashing algorithms don't need to match at all (to work) between peers' interconnected LAGs.

The most important thing is instead to use on both ends the LACP (IEEE 802.3ad) as the aggregation control protocol (it means that Port Trunks configured on HP should use the "lacp" option not the "trunk" option).

[tobys]

@parnassus Thanks very much!

this piece of info is very useful and not readily available or mentioned "Hashing algorithms don't need to match at all (to work) between peers' interconnected LAGs."