HP StorageWorks Fabric OS 6.3.1c Release Notes (5697-0509 September 2010 - includes all 6.3.1x versions)
Fixed in Fabric OS 6.3.1b
Switch panic or inability to execute any command may be experienced
after shared memory segments or memory shortage. It can be exposed
via:
• Creating and removing logical switch multiple times
• Performing configdownload with a large configuration file or repeat-
ing configdownload
• Running repeated perfresourceshow commands with option eeRes
or fportRes
• Launching port admin and selecting GbE port tab and keeping it
open for long time (2 hours)
• Cald being busy and not responding back to rpcd for some xml re-
quests causing stale shared memory segments
Fixed in Fabric OS 6.3.1b
If backbone and edge fabric have different switch PID formats, such as
when edge fabric is in IM2 mode while backbone fabric is not, import-
ing another device into the backbone will fail with [FCR-1022] "Local
proxy device slot entries exhausted" in raslog. Both routers and switches
configured for IR are impacted, and this issue is more likely to be seen
if there are frequent device offline/online transitions in edge fabrics.
Fixed in Fabric OS 6.3.1b
Workaround prior to upgrade: Once
switch is disabled, the configuration
downloaded and the switch
re-enabled, the configuration can be
modified and downloaded without
any problems. Can also remove or
comment out the link reset settings in
the .cfg file.
Unable to download Fabric Watch settings to the switch when mis-
matched names (LinkReset/LReset) are used in the configuration file.
DCFM unable to replicate Fabric Watch setting to switch and fails with
raslog error. Occurs only with linkreset setting in configuration files.
Fixed in Fabric OS 6.3.1b
Switch resets under very rare condition of lost heartbeat while active
CP is running Fabric OS 6.3.x and standby CP is running 6.2.x during
firmware download.
Fixed in Fabric OS 6.3.1b
Workaround prior to upgrade: Use
CLI or EFCM for non-TI/redirect
zones) in IM2/IM3.
Unexpected reboot and failover when attempting to push zone DB from
DCFM in IM2/IM3 fabrics.
Fixed in Fabric OS 6.3.1b
In a routed fabric environment (routers and/or switches running IR
mode), WWN for a device shows up in multiple edge fabrics. Proxy
devices for the same host/target could be created in unintended fabric
which could cause traffic to be sent to the wrong fabric and the host
not seeing the target. Occurs when there is frequent F port bounce (1-
2 sec) in FCR fabric. Issue exists only in Fabric OS 6.3.x.
Fixed in Fabric OS 6.3.1b
4Gb director does not respond to Link Reset primitives correctly. Due
to timing condition after reboot or switchdisable and switchenable, or
when device detects credit shortage, devices sent LR to the director but
it does not respond and the link goes down.
Fixed in Fabric OS 6.3.1b
Workaround prior to upgrade:
Disable POST on encryption switch
or chassis with encryption switch, or
reenable the AG ports manually.
Access Gateway NPIV port gets disabled as "initiator/target connected
on N_Port" after upstream encryption switch was rebooted, or encryption
blade executes a Power On Self Test. When bpportloopbacktest is run
as part of a POST on the encryption switch or encryption blade, loop
back mode caused the FLOGI frames from the AG ports on the AG
switch to loopback and further disable those ports.
88