Release Notes

Send documentation comments to [e-mail]
42
)Cisco MDS 9000 Family Release Notes for Cisco MDS NX-OS Release 5.0(1a)
OL-21012-01
Caveats
Symptom: If you enter the show dmm tech-support command after a DMM Job fails, DMM cores
and the module reloads.
Workaround: This issue is resolved.
• CSCtd59561
Symptom: SME hosts are unable to see the tape drives presented to them.
Workaround: This issue is resolved.
• CSCtd66543
Symptom: A memory leak in the Fabric-Device Management Interface (FDMI) process occurs
whenever a port to which the FDMI host is connected is flapped. If there are many port flaps, the
FDMI process may run out of memory and fail because memory is lost each time a port goes down.
Workaround: This issue is resolved.
• CSCtd78318
Symptom: Incorrect counting of zone members prevents the active zone database from being copied
to the full zone database.
This situation occurs when you attempt copy a zone database (active or backup) to the full zone
database. This action creates a new session, and the session database has as many members as the
number of members in the active (or backup) database. The session is not counting unique members
in the zone database. Instead it is counting repeated members (across different zones) as separate
members. Because of the repeated members, the number of zone members in the zoning database
exceeds 20,000, which is the maximum supported number of members.
Workaround: This issue is resolved.
• CSCtd96328
Symptom: The no system health module slot loopback failure-action command does not work as
expected. When an interface goes into hardware failure state because of congestion on the port, this
command should help the interface not go into a hardware failure state, but it does not.
Workaround: This issue is resolved.
• CSCte27874
Symptom: An MDS switch shows high CPU usage when viewed in Device Manager. If you enter
the show process cpu command, the output shows the following:
PID Runtime(ms) Invoked uSecs 1Sec Process
----- ---------------- -------- --------- ----- -----------
1675 2215685686 268836864 1565842753 94.3 vrrp-eng
The virtual router redundancy protocol (VRRP) engine is using more than 90% of the available CPU
time. This situation is caused by improper handling of VRRP-HA (authentication header) packets
on the supervisor.
Workaround: This issue is resolved.
• CSCtc65441
Symptom: When excessive traffic or errors occur on the mgmt0 port, an MDS 9124 switch may fail
and reload because of a watchdog timeout error.
Workaround: This issue is resolved.
• CSCtc68140