Cisco Field Notices
- The MDS 9000, Nexus 5000, Nexus 7000 and UCS 6100 families of products will be switching to a new block of Organizationally Unique Identifier (OUI). As a result of this change, customers may experience problems with Cisco Fabric Manager (FM). This Field Notice applies to units manufactured after February 09, 2010. The new OUIs will be 00-05-9b and 00-05-73. The first 3 bytes of the MAC address are the OUI. For information on determining whether a unit is affected, please see the Problem Symptom section. - The purpose of the filter assembly was to meet Telcordia (GR-1089) EMI requirements. The Telcordia specifications were later changed in 2006, making the capacitor unnecessary. The capacitors were removed from the design with the version -05 shelves which started shipping in October 2006, so the currently shipping shelves are not at risk. Up to 2% of shelves with incorrect capacitors may experience a capacitor failure during the normal life of the shelf. Customers with version -03 and -04 shelves may contact Cisco for a replacement as instructed below. As an alternative to complete shelf replacement, Cisco developed an in-service inspection program to aid in identifying incorrectly built power filter assemblies. Version -03 and -04 shelves may be inspected in the field to determine whether an incorrect capacitor is installed. The inspection is not traffic impacting, but it is strongly recommended that it be conducted during a maintenance window. After shelves in the field are inspected, they are labeled by the inspectors with a sticker, if the capacitor is correct. Please note: As the inspection process is manual and relies on a subjective color assessment to determine capacitor type, it is subject to error. Since the initial offering of the inspection program, there have been a small number of known manual inspection escapes, where incorrectly built power filter assemblies were misidentified as correctly built power filter assemblies, and subsequently failed. Analysis is underway on these errors to see if any further corrective action may be warranted. Given the possibility of escapes in the inspection process, Cisco extends the offer to replacement of any shelf covered by the inspection process at the customer's request. Customers who wish to schedule an inspection may contact Cisco as instructed below. - During high call volume in production, CTIOS may restart or assert with a minidump, failure to accept new client connections due to previous exceptions that may cause agent login failure, button enablement issues related agent state or call control. If CTIOS Server node did not restart automatically, it may need to be cycled via Cisco ICM Service Control to recover from this condition. Note: We have observed the problem to be more frequent, exposed under load with variable degree of agent counts and call volumes depending on customer environment and call flows. - During high call volume in production, CTIOS may restart or assert with a minidump, failure to accept new client connections due to previous exceptions that may cause agent login failure, button enablement issues related agent state or call control. If CTIOS Server node did not restart automatically, it may need to be cycled via Cisco ICM Service Control to recover from this condition. Note: We have observed the problem to be more frequent, exposed under load with variable degree of agent counts and call volumes depending on customer environment and call flows. - Cards remain in a constant state of reset (Red FAIL LED ON) due to a 1.5V DC/DC converter failure. This has been root caused to a specific component within the DC/DC converter design. - On August 8, 2005, the Energy Policy Act of 2005 (H.R.6.ENR), was signed into law. Section 110 of this Act modified the time change dates for Daylight Saving Time (DST) in the U.S. Beginning in March of 2008, DST will begin on the second Sunday of March and end the first Sunday of November. For 2008 and beyond, the daylight saving time period will be: 2:00 A.M. on the second Sunday in March to 2:00 A.M. on the first Sunday in November. - Executing any of the following commands on a MDS switch or collecting the same information via Fabric Manager may cause a kernel panic on the supervisor: - To be exposed to this issue, a server in the Products Affected section above must be running one of the affected application versions listed in the table found in the Workaround/Solution section. - -
|