Emulex 355 - Manual (Page 35)

E MULEX M ODEL 355 SAN S TORAGE S WITCH
U SER ’ S G UIDE C HAPTER 3: S WITCH M ANAGEMENT
E MULEX C ORPORATION 32
P ART N UMBER 00041411-002 R EV . B
Port Recovery
The following settings are available.
Setting Description
Bad Device Recovery When a port is already inserted into a zone, the port transforms F8
Failure notifications into F7 Initialization notifications. When this
occurs, the port is bypassed and F7 Initialization notifications are
allowed in the zone. Once the initialization is complete, the Bad Zone
Recovery Policy is operational and prevents a port that continues to
transmit F8 Failure notifications from inserting into the zone.
Note: If this policy is disabled while the Bad Zone Recovery policy is
enabled, a zone that does go down will still allow the Bad Zone
Recovery policy to reset the zone and allow ports to be
reinserted.
When enabled, this policy prevents devices that send F8 Failure
notifications from inserting into a zone. The ability to remove devices
that generate F8 Failure notifications automatically and
instantaneously guarantees continual system operation.
When disabled, this policy allows devices that send F8 Failure
notifications to insert into a zone and does not consider F8 Failure
notifications when determining whether to insert a device or not.
Clear on Stall In situations where the switch is operating in switching mode, some
devices may fall into an operating mode where the device has
opened a target but has not released the connection to the target.
When this policy is enabled, the switch can detect this condition and
automatically recover when this situation arises.
Bypass on No Activity The switch detects the amount of time a data stream has gone
without receiving a comma. The time setting is set to 100 (.001
seconds). When this policy is enabled, the switch bypasses the
disruptive port when the threshold is exceeded.
Bypass on Ordered Set Error Ordered Set (OS) errors are detected and counted for each individual
port. When this policy is enabled, a port is bypassed when its OS
count exceeds the threshold setting. The threshold setting is based
on the number of ordered set errors identified in 10 seconds.
Note: This threshold setting can be adjusted on the Web Manager’s
Advanced Functions Switch Thresholds page.
Bypass on CRC Error Cyclic Redundancy Check (CRC) errors are detected and counted for
each individual port. When this policy is enabled, a port is bypassed
when its CRC count exceeds the threshold setting. The threshold
setting is based on the number of CRC errors identified in 10
seconds.
User intervention is required to return the port into the zone.
Recovery methods include replacing the defective component,
cycling power to the device on the port, removing and reinserting the
bypassed port, or cycling power to the switch.
Note: This threshold setting can be adjusted on the Web Manager’s
Advanced Functions Switch Thresholds page.