HP 1606 HP B-series Fabric OS 6.4.1b Release Notes (5697-0886, March 2011-incl - Page 56

Table 9 Fabric OS 6.4.0 closed defects (continued) Closed defect summary Solution During reboot, HP StorageWorks D2D Backup System device sends the explicit logout; however, the switch does not clean up the route properly, resulting in the hosts being unable to detect the tape devices. Fixed in Fabric OS 6.4.0 In a rare case, a buffer leak in the rekey daemon causes an HP StorageWorks Fixed in Fabric OS 6.4.0 Encryption SAN Switch fault during rekey. A power cycle is required to recover. The SNMP manager gets incorrect MIB information from a switch running Fabric OS 6.1.x, due to a change in SwFwEPortUtil and Sw- FwEPortPktl between SW_v5_7 to SW_v5_8 MIB. Fixed in Fabric OS 6.4.0 After upgrade from Fabric OS 6.2.x to 6.3.0b/c and then downgrade back Fixed in Fabric OS 6.4.0 down to Fabric OS 6.2.0x, observe critical message "BKSW-5002 " swd_drv.c, line: 1323. The command fcoe loginshow displays unexpected values when an LP21000 or LP21002 CNA is moved between switches. Fixed in Fabric OS 6.4.0 HP StorageWorks Encryption SAN Switch does not rejoin Encryption Group Fixed in Fabric OS 6.4.0 after a reboot, resulting in a split Encryption Group. After reboot 2408 FCoE switch, monitored flood of internal RASLOG messages on FCOE/NPIV port. No other functional impact. Fixed in Fabric OS 6.4.0 When port 254 is used in directors in non-FICON setup, after multiple Fixed in Fabric OS 6.4.0 hafailovers, port 254 may lose its name server entry, and/or CP reports a synchronization error while overall system HA maintains sync. With either the 2408 FCoE Switch or the 10/24 FCoE Blade, when selecting Fixed in Fabric OS 6.4.0 FCoE Login in the Element Manager under switch admin, the error message FCoE Login Management is disabled in the switch. During power off/on switch test with HP StorageWorks DC SAN or DC04 Fixed in Fabric OS 6.4.0 SAN Directors, a panic rather than a blade fault, can occur when blades get too hot, or there is a bad temperature sensor, or spurious sensor reading. After hafailover, the portlog parameters debug and Portlog events Fixed in Fabric OS 6.4.0 enable are not persistent and go back to default, or setting in previous reboot, rather than the new setting before hafailover. On an HP StorageWorks DC SAN Director Switch Multiprotocol Extension Fixed in Fabric OS 6.4.0 Blade or 1606 Extension Switch, the portdisable xge0 CLI command does not disable the port, even though the RAS log and switchshow list it as disabled. FCIP tunnel bounce with impairment test. Fixed in Fabric OS 6.4.0 Workaround prior to upgrade: Identify and fix cause of TCP Network retransmissions. Reduce tunnel maximum data rate to level that does not experience packet loss or corruption on the network. The Configdownload command incorrectly removes license keys from new Fixed in Fabric OS 6.4.0 replacement switch. This applies only to Fabric OS 6.2.x code. When Access Gateway switch is connecting to a secondary port on an 8Gb Fixed in Fabric OS 6.4.0 or 4Gb 48-port blade, if the primary port goes down, some frames may be Workaround prior to upgrade: wrongly routed to the Access Gateway switch's secondary port, and AG will Do not put AG switch at the secondary respond to these frames with ACC rather than drop. port on 48-port blades. Misleading error when attempting to set port name field on port 255 (unsupported port). Fixed in Fabric OS 6.4.0 56