Specifications

Open Issues, Known Behaviors, and Resolved Issues
ExtremeXOS 12.5.4 Release Notes
76
PD4-506754505 When creating a dynamic egress ACL to deny all traffic, the traffic still hits the ARP reply
counter but not denying the packets.
PD4-450852442 When configuring an ACL and using the keyword denyAll in the create access-
list command, the action is not stopping the ARP reply packets.
PD4-614541490 VPLS traffic stops after configuring MAC limit-learning.
PD4-720906222 Performing a save and reboot on a DUT with 5,000 dynamic ACLs causes one module to
go into a failed state while the remaining modules stay in ACL sync even after one hour.
PD4-690958111 After running the unconfigure switch all command, an I/O module may stick in
the booting state, resulting in the switch continually rebooting.
PD4-631700490 When PIM and PIM snooping are enabled on the same VLAN, and if that VLAN happens
to be a PIM egress VLAN, traffic forwarding does not happen on the VLAN.
PD4-854675001,
PD4-860561781
A BlackDiamond 20808 switch with a single MM running ExtremeXOS 12.3.2 software
fails to upgrade during a firmware upgrade process. It also fails during the firmware
downgrade process. You may need to try and install the firmware multiple times.
PD4-552222911 DUT is hanging with a busy message (dots) after configuring 7,250 ingress dynamic
ACLs with conduit errors.
PD4-757707981 Proxy ARP does not work in a PVLAN.
PD4-526514731 On a BlackDiamond 20800 series switch, the system experiences a random packet drop
from the MM to the packet processor health check loopback interface.
PD4-730820249 If a route prefix in a BlackDiamond 20800 series switch is more than 24, only 100,000
routes are stored in the hardware. For the route prefixes less than 24, then 512,000
routes are stored in the hardware.
PD4-734160880,
PD4-697230006
Extreme Networks does not support the same (S,G) stream entering a switch through
multiple constituent VLANs of a private VLAN or VLAN aggregation. If a source moves
from one constituent VLAN to another, run the clear igmp snooping command.
PD4-1048824150 Deleting meters reports an error after an access list is configured and unconfigured
multiple times.
Workaround: Configure another meter.
PD4-1066604251 In dual MM systems, whenever a meter deletion results in a timeout, the show meter
command output still shows the meter, even though the error meter could not be
found is displayed.
Workaround: Create and use another meter.
PD4-1064653511 The error ........***** Process hal pid: 474 died with signal:11 is
displayed on an active MM when 1,000 policy files are configured on multiple egress
ports.
Workaround: Configurations exceeding the scaling limit are not supported.
PD4-1064653532 When configuring an access list with traffic queues, with a 10G port as one of the egress
port, traffic is lost when the corresponding 10G I/O module is hot swapped.
Workaround: Disable and re-enable the I/O module.
PD4-1056439342 I/O modules reboot when egress rate shaping is configured above 1,000 traffic queues
with all egress ports.
Workaround: Configure egress rate shaping with the proper scaling limits.
PD4-861903959 When the backup MM is not yet synced (because of running the run msm-failover
command or running diagnostics) and an MM performs an MM-failover, the command is
accepted, however, both MMs will then reboot.
Workaround: Make sure that run-time diagnostics command checks that both MMs
are in sync before running the diagnostics.
Table 38: Open Issues, Platform-Specific and Feature PDs (Continued)
PD Number Description