Specifications

Open Issues, Known Behaviors, and Resolved Issues
ExtremeXOS 12.3.4 Release Notes
106
PD4-594067381 During a firmware upgrade, one or more of the I/ O modules may get stuck in a booting
state during the TFTP download process. After the firmware upgrade is complete,
ExtremeXOS software reboots the switch and the I/O modules should become operational
within 5 to 10 minutes. If this does not occur, run the firmware upgrade again on those
I/O modules only.
PD4-678584893 An error message on the MM while upgrading firmware causes a communication break with
the uC resulting in an inability to see the uC version. This does not impact system
performance.
PD4-677367811 Some I/O modules in a fully populated chassis with mixed firmware may fail after a system
reboot
Workaround: Hot swap the failed modules.
PD4-676629197 After hot-swapping an MM-A or power cycling a BlackDiamond 20808 switch, the command
output of the show power budget command shows incorrect power data for MM-A or
MM-B.
PD4-672900303 When creating a VLAN in a user defined VR, an error relating to exceeding the maximum
buffer queue is seen.
<Crit:Kern.Emergency> MM-A: r.(1578) kioSendNeighbo ur() Exceeded
max r.(1578)
This does not impact system performance.
PD4-702644654 The system log displays warning messages when moving a cable from one port to another.
<Info:vlan.msgs.portLinkStateUp> MM-A: Port 6:15 link UP at speed 1
Gbps and full-duplex
<Warn:HAL.SysHC.PktPrcssrExcpt> MM-A: Serdes Group 1 exception in
packet
PD4-681012111 When adding a new rule to a policy file with a match condition of any, the new rule and the
existing rule are not hit by the ACL counters.
Workaround: Reboot the switch.
PD4-672070469 Power cycling a chassis or reseating an XFM-1 module causes the XFM-1 module to not
display when running the show version command. The following critical error is also
shown in the log.
<Crit:HAL.Fabric.Critical> MM-A: pioennerBringDownFE200 Bringing
Down FE200 on XBAR 4 as the command EXTRUC_FE_READ_CMD Failed,
Reason: UCPIPE_ERROR
This issue is intermittent and only occurs one out of 10 times.
Workaround: Reseat the XFM-1 module or reboot the chassis.
PD4-679154651,
PD4-681011831
When rebooting a BlackDiamond 20808 switch, MM-A gets stuck in INIT state.
PD4-679100154 During one out of 9 power cycle tests on a fully loaded chassis with 10GX modules, three
of the 10GX modules reboot twice before becoming operational. These modules stick in
TFTP boot and reboot again. The following error messages are seen on the MM console.
<Warn:DM.Warning> MM-A: Slot-3 FAILED (1) CardExec (state BOOTING)
timed out
<Warn:DM.Warning> MM-A: Slot-3 Timeout in state BOOTING Please see
attach log for detail
<Warn:DM.Warning> MM-B: Slot-3 FAILED (1) Error on Slot-3
PD4-676312796 When hot swapping multiple I/O modules while continuing to send ARP request traffic to a
10GX module, after all I/O modules boot up, slot 7 shows as failed on MM-B, but shows as
operational on MM-A.
PD4-585104810 Ports are not getting the line rate as expected when no weights are configured for a QoS
profile.
PD4-418743921 Booting a BlackDiamond 20808 switch with an XM-8XB module takes 4 minutes after
disabling and enabling the I/O slot.
Table 49: Platform-Specific and Feature PDs (Continued)