#!/usr/bin/bash

# check_351x plugin

# This nagios plugin does basic health checks on a Sun 3510 and 3511 disk arrays.
# It takes various parameters. Here's a description:

usage () {
	echo "Usage: $0 -h hostname -c community -d (3511|3510) -e x,x,x,x,x,x,x,x,x,x,x,y"
	echo "		IMPORTANT: You MUST read the comments of this plugin script."
}

#	check_351x [options]
#	-h	hostname
# 	-c	snmp community string (default: public)
#	-d 	devicetype (should be either '3510' or '3511')
#	-e	expected hdd status values (please read below)

while getopts h:c:d:e: option
do
        case $option in
        h)      host=$OPTARG;;
        c)      snmpstring=$OPTARG;;
        d)      device=$OPTARG;;
        e)      expect=$OPTARG;;
        ?)      usage; exit 3;;
        esac
done

if [ -z ${host} ] || [ -z $device ] || [ -z $expect ]; then
        usage
        exit 3
fi

exp=`echo $expect | tr -s ',' ' '`

if [ $device != "3511" ] && [ $device != "3510" ]; then
	echo "This plugin only supports Sun 3510 and 3511 disk arrays."
	echo "Sure you know what you're doin?"
	usage
	exit 3
fi

# Default snmp community string
if [ -z ${snmpstring} ]; then
	snmpstring=public
fi

# This plugin uses snmpwalk. It expects to find it in the constant specified here:
snmpwalk_bin=/usr/local/bin/snmpwalk

# OID to find hddStatus
hddoid=".1.3.6.1.4.1.42.2.180.$device.1.1.6.1.11"

# OID to find battery info
battoids=".1.3.6.1.4.1.42.2.180.$device.1.1.9.1.9.7 .1.3.6.1.4.1.42.2.180.$device.1.1.9.1.9.14"

# OID to find cachemodeflags
cacheoid=".1.3.6.1.4.1.42.2.180.$device.1.1.1.2.1.0"

# snmpwalk command to run for the hdd status check:

hdd_snmpwalk="${snmpwalk_bin} -m ALL -v 2c -O vq -c ${snmpstring} ${host} ${hddoid}"
hdd_response=`$hdd_snmpwalk | tr -s '\n' ' '`


if echo $hdd_response | grep -i "No Such Object" > /dev/null 2>&1; then
	echo "UNKNOWN: This device [$host] is not a Sun $device"
	exit 3
elif echo $hdd_response | grep -i "No more variables" > /dev/null >&1; then
	echo "UNKNOWN: This device [$host] is not a Sun $device"
	exit 3
fi

maxhdd=`echo $hdd_response | wc -w | tr -d ' 	'`

numdisks=`echo "$exp" | wc -w | tr -d ' 	'`

check_hdd_result () {

	if [ $1 -ge 128 ] && [ $1 -le 143 ]; then
		echo "SCSI Device"
		return 0
	fi

	val[0]="New Drive"
	val[1]="On-Line Drive"
	val[2]="Used Drive"
	val[3]="Spare Drive"
	val[4]="Drive Initializing"
	val[5]="Drive rebuilding"
	val[6]="Adding drive to logical drive"
	val[9]="Global Spare Drive"
	val[17]="Drive cloning"
	val[18]="Drive is valid clone"
	val[19]="Drive copying"
	val[63]="Drive absent"
	val[252]="Missing Global Spare Drive"
	val[253]="Missing Spare Drive"
	val[254]="Missing Drive"
	val[255]="Failed Drive"

	if [ $1 -ge 63 ]; then
		echo "Error $1: ${val[$1]}."
		return 2
	else
		echo "Result $1: ${val[$1]}."
		return 1
	fi
}

i=1
status=0

while [ $i -le $numdisks ]; do
	if [ $i -gt $maxhdd ]; then
		echo "CRITICAL: Ran out of disks. Either you specified too many, or there is a serious problem."
		exit 2
	fi	

	result=`echo "$hdd_response" | cut -f${i} -d' '`
	shouldbe=`echo "${expect}" | cut -f${i} -d','`
	if [ $result -ne "$shouldbe" ]; then
		output="$output Disk No:${i} `check_hdd_result $result`"
		if [ $? -eq 2 ]; then
			status=2
		elif [ $status -eq 0 ]; then
			status=1
		fi
	fi

	i=`expr $i + 1`
done

# snmpwalk command to run for the battery status check:

for oid in $battoids; do
	batt_snmpwalk="${snmpwalk_bin} -m ALL -v 2c -O vq -c ${snmpstring} ${host} ${oid}"
	batt_response="`$batt_snmpwalk`"
	tobinary=`echo "obase = 2 ; ${batt_response}" | bc`
	result=`printf "%08d" "$tobinary" | cut -c 1,2,5,8`
	if [ $result -gt 0 ]; then
		status=2
		output="$output Battery error."
	fi
done

# snmpwalk command to run for the cache status check:

cache_snmpwalk="${snmpwalk_bin} -m ALL -v 2c -O vq -c ${snmpstring} ${host} ${cacheoid}"
cache_response="`${cache_snmpwalk}`"
result=`expr $cache_response % 2`

if [ $result -eq 0 ]; then
	output="Cache disabled! $output"
	status=2
fi


# Final result processing

if [ $status -eq 2 ]; then
	echo "CRITICAL: $output"
	exit 2
elif [ $status -eq 1 ]; then
	echo "WARNING: $output"
	exit 1
elif [ $status -eq 0 ]; then
	echo "OK: Nothing out of the ordinary detected."
	exit 0
else
	echo "UNKNOWN: An unknown error has occurred"
	exit 3
fi

## HDD STATUS

# A bit about hdd status values:
# If you take a look at the MIB supplied by Sun, and do a bit of poking around,
# you'll see that the OID we're interested in for hdd status values is this:
#	.1.3.6.1.4.1.42.2.180.3511.1.1.6.1.11

# If you snmpwalk this OID, you'll see something like this:
# (Note: you'll need to copy the mib to /usr/local/share/snmp/mibs to see a proper description)

# 	# snmpwalk -m ALL -v 2c -c public se3511 .1.3.6.1.4.1.42.2.180.3511.1.1.6.1.11
#	SUN-STOREDGE-3511-MIB::hddStatus.1 = INTEGER: 1
#	SUN-STOREDGE-3511-MIB::hddStatus.2 = INTEGER: 1
#	SUN-STOREDGE-3511-MIB::hddStatus.3 = INTEGER: 1
#	SUN-STOREDGE-3511-MIB::hddStatus.4 = INTEGER: 1
#	SUN-STOREDGE-3511-MIB::hddStatus.5 = INTEGER: 1
#	SUN-STOREDGE-3511-MIB::hddStatus.6 = INTEGER: 1
#	SUN-STOREDGE-3511-MIB::hddStatus.7 = INTEGER: 1
#	SUN-STOREDGE-3511-MIB::hddStatus.8 = INTEGER: 1
#	SUN-STOREDGE-3511-MIB::hddStatus.9 = INTEGER: 1
#	SUN-STOREDGE-3511-MIB::hddStatus.10 = INTEGER: 1
#	SUN-STOREDGE-3511-MIB::hddStatus.11 = INTEGER: 1
#	SUN-STOREDGE-3511-MIB::hddStatus.12 = INTEGER: 9
#	SUN-STOREDGE-3511-MIB::hddStatus.13 = INTEGER: 141

# The hard drives are indexed 1 to 13, with 1 to 12 being actual hard drives
#	and 13 being the controller.

# The integer values have the following meanings:

# 0 : New Drive
# 1 : On-Line Drive
# 2 : Used Drive
# 3 : Spare Drive
# 4 : Drive Initialization in Progress
# 5 : Drive Rebuild in Progress
# 6 : Add Drive to Logical Drive in Progress
# 9 : Global Spare Drive
# 17 : Drive is in process of Cloning another Drive
# 18 : Drive is a valid Clone of another Drive
# 19 : Drive is in process of Copying from another Drive
# 	(for Copy/Replace LD Expansion function)
# 63 : Drive Absent
# 128 to 143 : SCSI Device
# 252 : Missing Global Spare Drive
# 253 : Missing Spare Drive
# 254 : Missing Drive
# 255 : Failed Drive"

# OK, so if we go back to our snmpwalk output, we can see that hdd's from 1 to 11
# are 'Online', hdd 12 is 'Global Spare Drive', and hdd 13 is in fact a SCSI device of some sort.

# Depending on your config, you will expect different status values, hence the -e parameter. 
# All you need to do is create a CSV list of the values you expect, and in what order. For example:
# 	-e '1,1,1,1,1,1,1,1,1,1,1,9'
# -- will do fine. This plugin will ignore the values of any unspecified hdd's (in this case, hdd 13 was 
# unspecified, and will be ignored).

# So, an example invokation:
# check_351x -h arrayhost -c public -d 3511 -e '1,1,1,1,1,1,1,1,1,1,1,9'


## BATTERY STATUS

# You also want to be informed of any failure with the battery.
# In the 3510 and 3511 that I played with, the two battery oid's were:
# .1.3.6.1.4.1.42.2.180.3511.1.1.9.1.9.7 and .1.3.6.1.4.1.42.2.180.3511.1.1.9.1.9.14
# (Yes, there are two batteries).
# If this is different, then change the battoids variable.

# According to the mib, the battery status is represented by converting the integer into a binary number
# then examining which bits are set to determine the state of the battery:

# BIT 0 - 	0:	Battery functioning normally.
#		1:	Battery malfunctioning.
# BIT 1 - 	0:	Battery charging OFF (or trickle).
#		1:	Battery charging ON.
# BIT 2 to 3 	If == 0, battery fully charged.
#		If == 1, battery not fully charged.
#		If == 2, battery charge critically low.
#		If == 3, battery completely drained.
# BIT 4 to 5  	Reserved (Set to 0).
# BIT 6 - 	0:	Battery-backup is enabled.
#		1:	Battery-backup is disabled.
# BIT 7 - 	0:	Battery IS present.
#		1:	Battery is NOT present.
# Value of 255 (ie, 1111 1111 or 0xff) means  "Status unknown".
#

# This plugin will return a CRITICAL if bit 0, 3, 6 or 7 is set. 
# Otherwise it will return an OK.

## CACHE STATUS
# MIB is .1.3.6.1.4.1.42.2.180.3511.1.1.1.2.1.0 mod 2 should be 1, else return CRIT "Caching disabled."