HP Cisco Nexus 5000 Cisco Nexus 5000 Series Release Notes Release 4.1(3)N1(1) - Page 18

Symptom, Workaround, Workaround, Title, vlan intf delete, system, qos service-policy, system qos

Page 18 highlights

Caveats Send documentation comments to [email protected] • CSCsw64952 Symptom: If the Cisco Nexus 5000 Series switch has more than 3K STP instances and an STP hello time of 1 second is used in the Bridge assurance configuration, the STP may not converge. Workaround: Use aa STP hello time of 2 second in the Bridge Assurance mode. • CSCsx67695 Symptom: When a gatos expansion module (GEM) port has an interface policy configuration that is consistent with the system policy, then when the GEM is removed from the switch, the system policy is changed to a new one. When the GEM is inserted into the switch, there can be an inconsistency between the GEM port interface policy and the new system policy (the new system policy may not have a system class that was in the old system policy). No error is reported for this inconsistency. Workaround: Remove the GEM port interface policy, fix the interface policy, then reapply to the GEM port. • CSCsx39481 Title: After vlan intf delete, its ipv6 address cannot be added to a diff. Symptom: After deleting an up vlan interface which contains an ipv6 address, you cannot add the same ipv6 address to another vlan interface. This is only seen with ipv6 addresses.IPV4 addresses do not have this problem. Workaround: Bring back the deleted VLAN interface, delete the IPv6 address and then delete the VLAN interface again. Followed by the addition of the same IPv6 address on a different VLAN interface. • CSCsx54086 Symptom: If source-vlan is configured for a monitor session and a downgrade is done from 4.0(1a)N2(1), the source-vlan configuration will be lost. The impact is the src-vlan traffic may not get spanned as desired after downgrade. Workaround: After downgrade, reconfigure the source-vlan configuration for a monitor session • CSCsx54270 Symptom: When upgrading to 4.0(1a)N2(1), if all cos values map to no-drop classes in the system qos service-policy configuration, the service-policy application will fail. The impact is the traffic may not conform to the expected behavior after upgrade. Workaround: After upgrade, reconfigure the policymap to include atleast one cos in a non no-drop class and reapply policymap to system qos service-policy configuration. • CSCsv52871 Symptom: When multiple FEX host ports congest FEX uplinks with no-drop class of traffic, pause asserted on the FEX host ports is uneven. As a result, traffic from some FEX host ports observe a better throughput than others. Workaround: None • CSCsx24526 Symptom: When a FEX host port is congested, an uncongested port in the same block of 8 ports will experience some traffic loss. The loss varies based on how many sources are trying to congest the congested port. Workaround: None Cisco Nexus 5000 Series and Cisco Nexus 2000 Series Release Notes, Release 4.1(3)N1(1) 18 OL-16601-01

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
  • 20
  • 21
  • 22
  • 23
  • 24
  • 25
  • 26

Send documentation comments to [email protected]
18
Cisco Nexus 5000 Series and Cisco Nexus 2000 Series Release Notes, Release 4.1(3)N1(1)
OL-16601-01
Caveats
CSCsw64952
Symptom
: If the Cisco Nexus 5000 Series switch has more than 3K STP instances and an STP hello
time of 1 second is used in the Bridge assurance configuration, the STP may not converge.
Workaround
: Use aa STP hello time of 2 second in the Bridge Assurance mode.
CSCsx67695
Symptom
: When a gatos expansion module (GEM) port has an interface policy configuration that
is consistent with the system policy, then when the GEM is removed from the switch, the system
policy is changed to a new one. When the GEM is inserted into the switch, there can be an
inconsistency between the GEM port interface policy and the new system policy (the new system
policy may not have a system class that was in the old system policy). No error is reported for this
inconsistency.
Workaround:
Remove the GEM port interface policy, fix the interface policy, then reapply to the
GEM port.
CSCsx39481
Title:
After
vlan intf delete
, its ipv6 address cannot be added to a diff.
Symptom
: After deleting an up vlan interface which contains an ipv6 address, you cannot add the
same ipv6 address to another vlan interface. This is only seen with ipv6 addresses.IPV4 addresses
do not have this problem.
Workaround:
Bring back the deleted VLAN interface, delete the IPv6 address and then delete the
VLAN interface again. Followed by the addition of the same IPv6 address on a different VLAN
interface.
CSCsx54086
Symptom
:
I
f source-vlan is configured for a monitor session and a downgrade is done from
4.0(1a)N2(1), the source-vlan configuration will be lost. The impact is the src-vlan traffic may not
get spanned as desired after downgrade
.
Workaround:
After downgrade, reconfigure the source-vlan configuration for a monitor session
CSCsx54270
Symptom
: When upgrading to 4.0(1a)N2(1), if all cos values map to no-drop classes in the
system
qos service-policy
configuration, the service-policy application will fail. The impact is the traffic
may not conform to the expected behavior after upgrade.
Workaround:
After upgrade, reconfigure the policymap to include atleast one cos in a non no-drop
class and reapply policymap to
system qos service-policy
configuration.
CSCsv52871
Symptom
: When multiple FEX host ports congest FEX uplinks with no-drop class of traffic, pause
asserted on the FEX host ports is uneven. As a result, traffic from some FEX host ports observe a
better throughput than others
.
Workaround:
None
CSCsx24526
Symptom
: When a FEX host port is congested, an uncongested port in the same block of 8 ports
will experience some traffic loss. The loss varies based on how many sources are trying to congest
the congested port.
Workaround:
None