Hi Community,

We have the next scenario.

2 x S5224F forming a VLT domain 2 in CPD-2  --> connected to other VLT domain 1 in CPD-1(equal switches)

1 x Qorestor server connected to VLT domain 2 with a static virtual port-channel (one cable to each switch).

 

On the virtual port-channel 21 ports, we see that traffic balancing is done correctly for inbound traffic but not for outbound traffic. Attach information of interfaces, load-balancing and hashing

S5224F-ON (1)

Ethernet 1/1/21:1 is up, line protocol is up
Description: Qorestor
Port is part of Port-channel 21
Hardware is Eth, address is 74:86:e2:75:92:ba
Current address is 74:86:e2:75:92:ba
Pluggable media present, SFP+ type is SFP+ 10GBASE-T
Wavelength is 0
Interface index is 39
Internet address is not set
Mode of IPv4 Address Assignment: not set
Interface IPv6 oper status: Disabled
MTU 9216 bytes, IP MTU 9184 bytes
LineSpeed 10G, Auto-Negotiation off
Flowcontrol rx off tx off
ARP type: ARPA, ARP Timeout: 60
Tag Protocol IDentifier (TPID) value: 0x8100
Last clearing of "show interface" counters: 2 weeks 4 days 23:58:48
Queuing strategy: fifo
Input statistics:
505158710 packets, 2306017791679 octets
232192683 64-byte pkts, 1784412 over 64-byte pkts, 15087500 over 127-byte pkts
167962 over 255-byte pkts, 89890 over 511-byte pkts, 255836263 over 1023-byte pkts
0 Multicasts, 11 Broadcasts, 505158699 Unicasts
0 runts, 0 giants, 0 throttles
0 CRC, 0 overrun, 0 discarded
Output statistics:
1954910822 packets, 13577078862102 octets
75083594 64-byte pkts, 351969759 over 64-byte pkts, 12144942 over 127-byte pkts
49918 over 255-byte pkts, 933574 over 511-byte pkts, 1514729035 over 1023-byte pkts
394091 Multicasts, 55488 Broadcasts, 1954461249 Unicasts
0 throttles, 6 discarded, 0 Collisions, wred drops
Rate Info(interval 30 seconds):
Input 410 Mbits/sec, 5782 packets/sec, 4% of line rate
Output 28 Mbits/sec, 10178 packets/sec, 0% of line rate
Time since last interface status change: 4 days 05:02:34

S5224F-ON (2)

Ethernet 1/1/21:1 is up, line protocol is up
Description: Qorestor
Port is part of Port-channel 21
Hardware is Eth, address is 74:86:e2:75:94:ba
Current address is 74:86:e2:75:94:ba
Pluggable media present, SFP+ type is SFP+ 10GBASE-T
Wavelength is 0
Interface index is 39
Internet address is not set
Mode of IPv4 Address Assignment: not set
Interface IPv6 oper status: Disabled
MTU 9216 bytes, IP MTU 9184 bytes
LineSpeed 10G, Auto-Negotiation off
Flowcontrol rx off tx off
ARP type: ARPA, ARP Timeout: 60
Tag Protocol IDentifier (TPID) value: 0x8100
Last clearing of "show interface" counters: 2 weeks 4 days 20:47:39
Queuing strategy: fifo
Input statistics:
502492133 packets, 2309381492941 octets
228994932 64-byte pkts, 2076808 over 64-byte pkts, 14909474 over 127-byte pkts
167452 over 255-byte pkts, 102419 over 511-byte pkts, 256241048 over 1023-byte pkts
0 Multicasts, 2 Broadcasts, 502492131 Unicasts
0 runts, 0 giants, 0 throttles
0 CRC, 0 overrun, 0 discarded
Output statistics:
12214 packets, 4673373 octets
9 64-byte pkts, 6 over 64-byte pkts, 0 over 127-byte pkts
12199 over 255-byte pkts, 0 over 511-byte pkts, 0 over 1023-byte pkts
12211 Multicasts, 0 Broadcasts, 3 Unicasts
0 throttles, 0 discarded, 0 Collisions, wred drops
Rate Info(interval 30 seconds):
Input 304 Mbits/sec, 4284 packets/sec, 3% of line rate
Output 0 Mbits/sec, 0 packets/sec, 0% of line rate
Time since last interface status change: 4 days 05:04:38

 

Load-Balancing Configuration For LAG and ECMP:
----------------------------------------------
IPV4 Load Balancing : Enabled
IPV6 Load Balancing : Enabled
MAC Load Balancing : Enabled
TCP-UDP Load Balancing : Enabled
Ingress Port Load Balancing : Enabled
IPV4 FIELDS : source-ip destination-ip protocol vlan-id l4-destination-port l4-source-port
IPV6 FIELDS : source-ip destination-ip protocol vlan-id l4-destination-port l4-source-port
MAC FIELDS : source-mac destination-mac ethertype vlan-id
TCP-UDP FIELDS : l4-destination-port l4-source-port

dell#show hash-algorithm
LagAlgo - CRC EcmpAlgo - CRC
dell#

Reviewing Dell documentation, I see that load-balancing and hashing can be configured but at the global switch level, can the changes made affect the rest of the virtual port-channels that I have connected in the VLT domain?

Any idea how to perform troubleshooting?

Thanks in advance!