HP 4400 HP StorageWorks Fabric OS 6.4.0c Release Notes (5697-0703, September 2 - Page 54
Fixed in Fabric OS 6.4.0. New
View all HP 4400 manuals
Add to My Manuals
Save this manual to your list of manuals |
Page 54 highlights
Closed defect summary Solution Zoning restriction does not work for NPIV/loop ports when one device tries to get another devices information on the same port using name based queries (GID_PN, GID_NN), which could lead to device communication failures. Fixed in Fabric OS 6.4.0 Host failed to discover tapes due to a VI/VT (Virtual Initiator/Virtual Target) database mismatch between the switches that are part of the same Encryption Group (EG). Therefore, the host cannot discover tape devices that are part of the same EG. Fixed in Fabric OS 6.4.0 Workaround prior to upgrade: Issue "cryptocfg --commit" again to sync up the VT/VI database across nodes in the EG. In FICON environment, detected termination of process ficud:2658, running 6.2.0e. Fixed in Fabric OS 6.4.0 Ports on the DC SAN director 48 port 8Gb blade become persistently disabled after upgrading from Fabric OS 6.2.x to Fabric OS 6.3.x with VF disabled and fmsmode enabled. Fixed in Fabric OS 6.4.0 Workaround prior to upgrade: Manually re-enable ports. No other viable workaround at this time. After moving some servers from backbone to edge, FCR-1022 messages are seen and devices are no longer getting exported into backbone. If backbone and edge fabric have different switch PID formats such as when edge fabric is in IM2 mode, while backbone fabric is not, unimport an device did not free up resource. When resource is exhausted, import another device into backbone will fail with [FCR-1022] " Local proxy device slot entries exhausted" in raslog. Both 4Gb and 8Gb FCRs are impacted and are more likely to be seen if there is frequent device offline/online in edge fabrics. Fixed in Fabric OS 6.4.0 Termination of non-restartable daemons always triggers failover/reboot for recovery, Termination of non-VF and non-FSS compliant daemons (raslogd and traced) always trigger stateless restart. However, in certain cases, cold recovery can be triggered by the restart of a restartable daemon. Fixed in Fabric OS 6.4.0 Enhancement request to add a user selectable parameter to adjust SFP polling during high CPU load situation. Fixed in Fabric OS 6.4.0. New parameter added to Fabric OS 6.4.0 to allow adjustment of SFP polling interval to avoid laser faults. Invalid zone name does not cause error message, and no enforcement Fixed in Fabric OS 6.4.0 of 64 character limit on cfg/zone name. Functional impact is zone is not enabled. Need additional information of EE (Encryption Engine) status for the entire EG (Encryption Group) in a single command. Currently we have to invoke cryptocfg --show -groupcfg", cryptocfg --show -groupmember -all, or cryptocfg --show -localEE on all nodes to know the overall state of EG including EE state across the EG. Fixed in Fabric OS 6.4.0 When auditing is on for security class, a race condition can occur in raslogd, which leads to a switch panic or hafailover. Fixed in Fabric OS 6.4.0 Workaround prior to upgrade: Disable auditing via: Auditcfg --disable 54