cled(7) cled(7)
NAME
cled - Compaq ProLiant Storage System Driver
DESCRIPTION
This driver supports the Compaq ProLiant Storage System when
connected to a Compaq 32-Bit Fast-SCSI-2 Controller using the
cpqsc host adapter driver. All models of the Compaq ProLiant
External Storage System and the Compaq ProLiant 1000 and
ProLiant 2000 Servers supporting Compaq Pluggable SCSI drives
are supported. This driver is included in the compaq package
for UnixWare. The component cpqscsimon is included in the
same package.
The functionality implemented includes support for the LED
Indicators on the Drive Tray for Compaq Pluggable SCSI drives
when installed in the ProLiant cabinet, status change
reporting for pluggable drives and for the ProLiant cabinet
via console messages and UnixWare mail, and status change
reporting via SNMP when Compaq Insight Manager Agents for
UnixWare are installed. The support is installable only on
EISA based systems.
UnixWare installation will automatically configure this
support on all systems with Compaq 32-Bit Fast-SCSI-2
Controllers supported by the cpqsc driver when the compaq
package is added.
ioctl Calls
The ioctl calls used by the cled driver to report on the
condition of the ProLiant Storage System are as follows:
SDI_IOC_CLED_GET_HA_INDICATORS
Used to return the status of the drive slots. The len member
of the CLED_REQUEST structure shown below must point to a
buffer large enough for an 8-entry (one per SCSI target id)
array of the following structure:
struct cled_id {
time_t online_since; /* time at last online status change */
time_t service_since; /* time at last service status change */
uchar_t online_status; /* present online status */
uchar_t service_status; /* present service status */
ushort_t flags; /* per device flags */
};
typedef struct cled_id CLED_ID;
Copyright 1994 Novell, Inc. Page 1
cled(7) cled(7)
SDI_IOC_CLED_GET_HA_INDICATORS
Used to return the global information maintained for each host
adapter/ProLiant Storage System combination. The structure
member len in CLED_REQUEST must point to a buffer large enough
to contain the following structure, returned by the driver:
struct cled_global {
char name[8]; /* host adapter name */
uchar_t ha; /* host adapter number */
uchar_t reserved[3];
uint_t flags; /* ProLiant cabinet (nile) status flags */
char vendor[9]; /* ProLiant Vendor, trailing null */
char product[17]; /* ProLiant Product, trailing null */
char version[5]; /* ProLiant ROM Version, trailing null */
short fan_status; /* ProLiant Fan Status */
short reserved2; /* future status */
short temp_status; /* ProLiant Temperature Status */
short reserved3; /* future status */
short door_status; /* ProLiant Door Status */
short power_status; /* ProLiant Redundant Power Status */
ulong_t slots_present; /* ProLiant slots present mask */
ulong_t drives_present; /* ProLiant drives present mask */
};
typedef struct cled_global CLED_GLOBAL;
The argument to the ioctls is the address of the following
structure:
struct cled_request {
char name[8]; /* host adapter name */
uchar_t ha; /* host adapter number */
uchar_t id; /* scsi id (if used) */
ushort_t len; /* length of info structure buffer */
void *p; /* pointer to info structure buffer */
};
typedef struct cled_request CLED_REQUEST;
ProLiant Storage System Messages
The ProLiant Storage System driver, cled, tracks the condition
of the ProLiant Storage System and issues messages to the
system console if it finds any failures or performance
degradation. When in multiuser mode, system console messages
are normally also recorded in the file /usr/adm/messages.
Copyright 1994 Novell, Inc. Page 2
cled(7) cled(7)
The following list gives definitions of parameters used by the
cled driver in many of the following messages:
ha the unique host adapter number which was set during
System Configuration for each 32-Bit Fast-SCSI-2
Controller in your system. This is the number UnixWare
uses to distinguish among multiple controllers of the
same type. The first SCSI controller of each type is
number 0.
id the unique SCSI ID of a peripheral connected to a SCSI
bus. The id displayed identifies a particular SCSI
peripheral on a SCSI bus associated with this message.
When id is displayed, ha is also displayed to identify
the SCSI bus.
NOTE: If you are unsure which SCSI peripheral is
configured on which bus or at which id, the utility
cpqsmu displays peripherals discovered on the SCSI bus.
Also, the PDI utility pdiconfig will show all SCSI
peripherals configured into the kernel.
ht is the host adapter type. It is displayed as an ASCII
string. The string contains the name of the host
adapter driver for the host adapter controlling this
SCSI bus. For example, ht=cpqsc. More than one type of
SCSI host adapter can be present in a single system.
Initialization Messages
cled: ProLiant Temperature Status is ALARM ht=ht ha=ha
Drive auto spin down has been enabled
Make sure all covers are closed
Make sure all fans are operating
Make sure air flow is unobstructed
Check room temperature (cledW01)
The message above indicates that the temperature in a ProLiant
drive enclosure is too high. This condition should be
corrected as quickly as possible to protect the drives from
severe damage including drive failure and resulting data loss.
cled: ProLiant Temperature Status is CRITICAL ht=ht ha=ha
Power the unit off to allow the drives to cool
Make sure all covers are closed
Make sure all fans are operating
Make sure air flow is unobstructed
Copyright 1994 Novell, Inc. Page 3
cled(7) cled(7)
Check room temperature (cledW02)
The message above indicates that the temperature in a ProLiant
drive enclosure is extremely hot and drive damage is highly
likely. Take immediate action to shut down and cool these
drives.
cled: ProLiant Fan Status is ALARM ht=ht ha=ha
Drive auto spin down has been enabled
Power the unit off to allow the drives to cool
Check and repair the fan (cledW03)
The message above indicates that a monitored fan in a ProLiant
drive enclosure has failed. Locate the failed fan and repair
this condition as soon as possible to protect the drives from
severe damage including drive failure and resulting data loss.
cled: ProLiant Door Status is ALARM ht=ht ha=ha
Make sure all covers are closed (cledW04)
This message indicates that a monitored door in a ProLiant
drive enclosure has been opened. Correct air flow to maintain
proper cooling of the drives is not possible with the cabinet
cover removed. Replace the cover as soon as possible to
protect the drives from severe damage including drive failure
and resulting data loss.
cled: Unable to allocate memory at init time (x bytes)
The cled driver has failed to initialize
ProLiant features are not available (cledW05)
where x is the amount of memory, in bytes, that was requested.
The message above indicates that a request for memory at
initialization time has been refused by the kernel. The
kernel you are booting is too large to run correctly with the
amount of memory detected by the kernel. Be sure your system
has sufficient RAM memory and that this memory is correctly
configured so that the kernel can detect it.
cled: ProLiant is OPERATIONAL ht=ht ha=ha
A ProLiant hard drive cabinet reported power up or reset
A ProLiant has been connected, powered on, or reset (cledW06)
This message indicates that a ProLiant Storage System which
was previously undetected or not operational has been restored
Copyright 1994 Novell, Inc. Page 4
cled(7) cled(7)
to operation.
cled: ProLiant is NOT OPERATIONAL ht=ht ha=ha
A ProLiant hard drive cabinet has stopped responding
The ProLiant is disconnected or not powered (cledW07)
This message indicates that a ProLiant Storage System which
was previously detected or operational is no longer
operational. The drives in the cabinet may also be
unavailable. Take steps to correct this condition. If any
drives in the ProLiant cabinet were online, data loss is
likely.
Notices
The following messages are preceded by the word NOTICE on the
system console. These messages may contain important
information; however, no immediate user action may be
required.
cled: ProLiant Temperature Status is OK ht=ht ha=ha (cledN02)
This message indicates that the temperature in a ProLiant
drive enclosure has changed to OK.
cled: ProLiant Fan Status is OK ht=ht ha=ha (cledN03)
This message indicates that the status of a monitored fan in a
ProLiant drive enclosure is now OK.
cled: ProLiant Door Status is OK ht=ht ha=ha (cledN04)
This message indicates that a monitored door in a ProLiant
drive enclosure has been closed.
cled: A ProLiant drive has been installed ht=ht ha=ha (cledN05)
This message indicates that a hot pluggable drive has been
installed in a ProLiant drive enclosure. If an online drive
was removed from this location, the newly installed drive is
spun up. Spinning up the drive will sometimes prevent data
loss when an online drive is removed by mistake.
cled: A ProLiant drive has been removed ht=ht ha=ha (cledN06)
Copyright 1994 Novell, Inc. Page 5
cled(7) cled(7)
This message indicates that a hot pluggable drive has been
removed from a ProLiant drive enclosure. If the drive was
intentionally removed, no further action is required. If this
drive was not offline, data loss is possible.
NOTE: Removal of hot pluggable drives is safe only when the
drive is offline as indicated by the Online LED on the drive
tray being off.
FILES
/etc/conf/pack.d/cled/*
/dev/cled/*
REFERENCES
cpqsc(7), cpqscsimon(1M), cpqsmu(1M)
NOTICES
This driver is only supported on applicable Compaq systems.
Copyright 1994 Novell, Inc. Page 6