Bug ID 519510: Throughput drop and rxbadsum stat increase in tagged VLAN with LRO/GRO on BIG-IP VE running on ESX platforms with particular network hardware

Last Modified: Nov 07, 2022

Bug Tracker

Affected Product:  See more info
BIG-IP LTM(all modules)

Known Affected Versions:
11.4.0, 11.4.1, 11.5.1, 11.5.1 HF1, 11.5.1 HF10, 11.5.1 HF11, 11.5.1 HF2, 11.5.1 HF3, 11.5.1 HF4, 11.5.1 HF5, 11.5.1 HF6, 11.5.1 HF7, 11.5.1 HF8, 11.5.1 HF9, 11.5.10, 11.5.2, 11.5.2 HF1, 11.5.3, 11.5.3 HF1, 11.5.3 HF2, 11.6.0, 11.6.0 HF1, 11.6.0 HF2, 11.6.0 HF3, 11.6.0 HF4, 11.6.0 HF5

Fixed In:
12.0.0, 11.6.0 HF6, 11.5.4, 11.4.1 HF10

Opened: Apr 22, 2015
Severity: 3-Major
Related Article:
K17164

Symptoms

TCP throughput might be severely impacted for traffic traversing a tagged VLAN and BCM57800/BCM57810 NIC on BIG-IP VEs. The 'rxbadsum' counts increase as received LRO'd traffic is ignored by TMM.

Impact

Potential throughput drop during a high volume of data transfer.

Conditions

1. Traffic traverses a tagged VLAN. 2. This issue might be related to systems using Broadcom BCM57800 or BCM57810 NICs. However in general, the required condition is reception of packets with VLAN header are received in uNIC driver.

Workaround

You can use either of the following workarounds: 1. Avoid using tagged VLANs. 2. Run the following commands on the ESX hypervisor to disable LRO/GRO system-wide, followed by a reboot. -- esxcli system settings advanced set -o /Net/Vmxnet2HwLRO -i 0. -- esxcli system settings advanced set -o /Net/Vmxnet3HwLRO -i 0. -- esxcli system settings advanced set -o /Net/Vmxnet2SwLRO -i 0. -- esxcli system settings advanced set -o /Net/Vmxnet3SwLRO -i 0. -- esxcli system settings advanced set -o /Net/VmxnetSwLROSL -i 0.

Fix Information

Change in L4 packet header offset, resulting from VLAN header insertion, is being accounted for to verify checksum.

Behavior Change