HP 1606 HP B-series Fabric OS 6.4.1b Release Notes (5697-0886, March 2011-incl - Page 49

Table 9 Fabric OS 6.4.0 closed defects (continued) Closed defect summary If Fabric Watch fwalarms is turned off and Errlog is turned on, fabric watch daemon can crash causing a switch panic. Solution Fixed in Fabric OS 6.4.0 Workaround prior to upgrade: In the threshold alarm matrix configuration for all port related thresholds, disable ErrLog and Port Fencing (using fwconfigure) this will avoid the above functions from being called (on any threshold events) and leaking or enable fwalarmsfilterset. EX_Port in DC04 SAN Director base switch in logical switch configuration fails to come up online as EX_Port. Fixed in Fabric OS 6.4.0 Workaround prior to upgrade: Always add FC ports before adding GE ports to an LS. When removing ports from an LS, remove GE ports before FC ports. On HP StorageWorks 1606 Extension SAN switch, no NPIV devices (including Access Gateway) can log into ports 0 - 3 of multiprotocol router blade or 400 MP router. Error message occurs stating NPIV not supported on Fabric port, although NPIV is enabled on the 1606 ports. No problem logging into ports 4 - 31. Fixed in Fabric OS 6.4.0 Port-to-port connectivity fails due to slow RNID processing on 4Gb switches. Fixed in Fabric OS 6.4.0 Host failed to discover tapes due to a VI/VT (Virtual Initiator/Virtual Target) Fixed in Fabric OS 6.4.0 database mismatch between the switches that are part of the same Encryption Workaround prior to upgrade: Group. Therefore, the host cannot discover tape devices that are part of the same EG. Issue cryptocfg --commit again to sync up the VT/VI database across nodes in the EG. Common Access Library Daemon (CALD) detects a large memory increase on the HP StorageWorks 2408 FCoE switch when running multiple SMIA clients over FCoE. Other Fabric OS-based products do not exhibit the same behavior when running the same scripts. Fixed in Fabric OS 6.4.0 HBA is seen in switchshow but not nsshow. When plugged two ports into Fixed in Fabric OS 6.4.0 a switch, both ports show logged in and online in switchshow, but the HBA device is missing from nsshow because the FLOGI response did not update PID properly. When small read requests (low I/O rate traffic) sent via FCIP tunnel, a 2 - 4 ms delay occurs between the time data arrives on the GE port until it is sent out from the FC port, leading to >10 ms response time. Fixed in Fabric OS 6.4.0 Workaround prior to upgrade: Do not run low I/O rate traffic.This occurs only when low I/O rate traffic is passing through the FCIP Tunnel. With a 2408 FCoE switch in Access Gateway mode, traffic stops on all host Fixed in Fabric OS 6.4.0 ports within a trunk when one of the hosts is disabled using the fcoe --disable command, or shutting that interface, or by pulling the cable on it. After upgrade to Fabric OS 6.2.x, the 400 MP router, or multiprotocol router Fixed in Fabric OS 6.4.0 blade observes a credit problem on virtual channel 0, leading to frame drop Workaround prior to upgrade: on FCIP and FC FastWrite, and the host being unable to see the target. Power On Reset of the router; or slotpower off/on of router blade, or Power On Reset the chassis. After a target port bounce, the FC FastWrite host cannot recognize the target Fixed in Fabric OS 6.4.0 due to cam setup problem on the backend port used by FC FastWrite. Fabric OS 6.4.0 fixes 49