| 1 | # you can disable an alarm notification by setting the 'to' line to: silent |
| 2 | |
| 3 | # ----------------------------------------------------------------------------- |
| 4 | # low disk space |
| 5 | |
| 6 | # checking the latest collected values |
| 7 | # raise an alarm if the disk is low on |
| 8 | # available disk space |
| 9 | |
| 10 | template: disk_space_usage |
| 11 | on: disk.space |
| 12 | class: Utilization |
| 13 | type: System |
| 14 | component: Disk |
| 15 | chart labels: mount_point=!/dev !/dev/* !/run !/run/* !HarddiskVolume* * |
| 16 | calc: $used * 100 / ($avail + $used) |
| 17 | units: % |
| 18 | every: 1m |
| 19 | warn: $this > (($status >= $WARNING ) ? (80) : (90)) |
| 20 | crit: ($this > (($status == $CRITICAL) ? (90) : (98))) && $avail < 5 |
| 21 | delay: up 1m down 15m multiplier 1.5 max 1h |
| 22 | summary: Disk ${label:mount_point} space usage |
| 23 | info: Total space utilization of disk ${label:mount_point} |
| 24 | to: sysadmin |
| 25 | |
| 26 | template: disk_inode_usage |
| 27 | on: disk.inodes |
| 28 | class: Utilization |
| 29 | type: System |
| 30 | component: Disk |
| 31 | chart labels: mount_point=!/dev !/dev/* !/run !/run/* * |
| 32 | calc: $used * 100 / ($avail + $used) |
| 33 | units: % |
| 34 | every: 1m |
| 35 | warn: $this > (($status >= $WARNING) ? (80) : (90)) |
| 36 | crit: $this > (($status == $CRITICAL) ? (90) : (98)) |
| 37 | delay: up 1m down 15m multiplier 1.5 max 1h |
| 38 | summary: Disk ${label:mount_point} inode usage |
| 39 | info: Total inode utilization of disk ${label:mount_point} |
| 40 | to: sysadmin |
| 41 | |
| 42 | |
| 43 | # ----------------------------------------------------------------------------- |
| 44 | # disk fill rate |
| 45 | |
| 46 | # calculate the rate the disk fills |
| 47 | # use as base, the available space change |
| 48 | # during the last hour |
| 49 | |
| 50 | # this is just a calculation - it has no alarm |
| 51 | # we will use it in the next template to find |
| 52 | # the hours remaining |
| 53 | |
| 54 | template: disk_fill_rate |
| 55 | on: disk.space |
| 56 | lookup: min -10m at -50m unaligned of avail |
| 57 | calc: ($this - $avail) / (($now - $after) / 3600) |
| 58 | every: 1m |
| 59 | units: GB/hour |
| 60 | info: average rate the disk fills up (positive), or frees up (negative) space, for the last hour |
| 61 | |
| 62 | # calculate the hours remaining |
| 63 | # if the disk continues to fill in this rate |
| 64 | |
| 65 | template: out_of_disk_space_time |
| 66 | on: disk.space |
| 67 | calc: ($disk_fill_rate > 0) ? ($avail / $disk_fill_rate) : (inf) |
| 68 | units: hours |
| 69 | every: 10s |
| 70 | warn: $this > 0 and $this < (($status >= $WARNING) ? (48) : (8)) |
| 71 | crit: $this > 0 and $this < (($status == $CRITICAL) ? (24) : (2)) |
| 72 | delay: down 15m multiplier 1.2 max 1h |
| 73 | summary: Disk ${label:mount_point} estimation of lack of space |
| 74 | info: Estimated time the disk ${label:mount_point} will run out of space, if the system continues to add data with the rate of the last hour |
| 75 | to: silent |
| 76 | |
| 77 | |
| 78 | # ----------------------------------------------------------------------------- |
| 79 | # disk inode fill rate |
| 80 | |
| 81 | # calculate the rate the disk inodes are allocated |
| 82 | # use as base, the available inodes change |
| 83 | # during the last hour |
| 84 | |
| 85 | # this is just a calculation - it has no alarm |
| 86 | # we will use it in the next template to find |
| 87 | # the hours remaining |
| 88 | |
| 89 | template: disk_inode_rate |
| 90 | on: disk.inodes |
| 91 | lookup: min -10m at -50m unaligned of avail |
| 92 | calc: ($this - $avail) / (($now - $after) / 3600) |
| 93 | every: 1m |
| 94 | units: inodes/hour |
| 95 | info: average rate at which disk inodes are allocated (positive), or freed (negative), for the last hour |
| 96 | |
| 97 | # calculate the hours remaining |
| 98 | # if the disk inodes are allocated |
| 99 | # in this rate |
| 100 | |
| 101 | template: out_of_disk_inodes_time |
| 102 | on: disk.inodes |
| 103 | calc: ($disk_inode_rate > 0) ? ($avail / $disk_inode_rate) : (inf) |
| 104 | units: hours |
| 105 | every: 10s |
| 106 | warn: $this > 0 and $this < (($status >= $WARNING) ? (48) : (8)) |
| 107 | crit: $this > 0 and $this < (($status == $CRITICAL) ? (24) : (2)) |
| 108 | delay: down 15m multiplier 1.2 max 1h |
| 109 | summary: Disk ${label:mount_point} estimation of lack of inodes |
| 110 | info: Estimated time the disk ${label:mount_point} will run out of inodes, if the system continues to allocate inodes with the rate of the last hour |
| 111 | to: silent |
| 112 | |
| 113 | |
| 114 | # ----------------------------------------------------------------------------- |
| 115 | # disk congestion |
| 116 | |
| 117 | # raise an alarm if the disk is congested |
| 118 | # by calculating the average disk utilization |
| 119 | # for the last 10 minutes |
| 120 | |
| 121 | template: 10min_disk_utilization |
| 122 | on: disk.util |
| 123 | class: Utilization |
| 124 | type: System |
| 125 | component: Disk |
| 126 | lookup: average -10m unaligned |
| 127 | units: % |
| 128 | every: 1m |
| 129 | warn: $this > 98 * (($status >= $WARNING) ? (0.7) : (1)) |
| 130 | delay: down 15m multiplier 1.2 max 1h |
| 131 | summary: Disk ${label:device} utilization |
| 132 | info: Average percentage of time ${label:device} disk was busy over the last 10 minutes |
| 133 | to: silent |
| 134 | |
| 135 | |
| 136 | # raise an alarm if the disk backlog |
| 137 | # is above 1000ms (1s) per second |
| 138 | # for 10 minutes |
| 139 | # (i.e. the disk cannot catch up) |
| 140 | |
| 141 | template: 10min_disk_backlog |
| 142 | on: disk.backlog |
| 143 | class: Latency |
| 144 | type: System |
| 145 | component: Disk |
| 146 | lookup: average -10m unaligned |
| 147 | units: ms |
| 148 | every: 1m |
| 149 | warn: $this > 5000 * (($status >= $WARNING) ? (0.7) : (1)) |
| 150 | delay: down 15m multiplier 1.2 max 1h |
| 151 | summary: Disk ${label:device} backlog |
| 152 | info: Average backlog size of the ${label:device} disk over the last 10 minutes |
| 153 | to: silent |