Nagios Plugins for NetApp: dependency table check_netapp

Configuration • Dependencies • The Matrix

About Dependencies

Most of the check_netapp_pro check-scripts get their data from a local store-file, which is written by a collector (also known as getter due to their file-name get_netapp*.pl). The getter-scripts are configured in the monitoring-system and run as if they were normal service-checks.

Some very simple checks do not require any collector. They run like a traditional stand-alone plugin and retrieve their data directly from the filer.

Please keep in mind that all checks from the Performance-Bundle get their data trough a special performance-data getter (get_netapp_perfdata.pl).

Examples

Some examples which show how the data for specific checks gets collected. The relation between the check and the object is fully documented in the table below.

Example 1: Single Dependency

The following example shows which getter must have been run so that the Usage of volumes can be checked on a cluster-mode filer:

./get_netapp_cm.pl -H my_cdot_filer --object=volume

Same as above but for a 7-mode filer:

./get_netapp_7m.pl -H my_7m_filer --object=volume

Example 2: Performance Data

To check the per volume latency (PerfVolume) the performance-getter must have been run (at least twice) configured with the correct object and mode as shown below:

./get_netapp_perfdata.pl -H my_cdot_filer --object=volume --mode=cm

Example 3: Complex Dependency³

Some checks need the data from more than one object. The following example shows the requirements to monitor the overcommitment of aggregates with the check OvercommitAggr:

./get_netapp_cm.pl -H my_cdot_filer --object=volume
./get_netapp_cm.pl -H my_cdot_filer --object=aggregate

The next example shows the requirements for checking Snapshots. Here we also see that the objects data must be either retrieved by the generic getter get_netapp_cm.pl or in some cases with a specialized getter for that object like get_netapp_snapshot.pl.

./get_netapp_cm.pl -H my_cdot_filer --object=volume
./get_netapp_snapshot.pl -H my_cdot_filer # the snapshot-object has its own getter

The Matrix of Dependencies by Bundle and Check

This table lists all dependencies between checks and getters. If no dependency is listed the check is a traditional stand-alone check.

Advanced Bundle

Check	Status		depends on object(s)
Check	7m	cm	depends on object(s)
AggregateState checks the aggregates-state. Alarms if they are not online (configurable). `[help]`	stable	stable	aggregate
AutosizeMode checks the autosize-mode of autosized volumes if they are all set to given value (grow, grow_shrink, ...) `[help]`	-	beta	volume
check_netapp7_cluster checks the status of the high availability service (connected, taken over, takeover failed, ...).	stable	-
check_netapp7_fcpstats monitors the FCP adapters for crc-errors and other values.	stable	-
check_netapp7_snapvault monitors the status and lag-time of Snapvault relations onyl on 7m filers. (Cdot filers are checked with the SnapMirror checks).	stable	-
check_netapp7_vfiler monitors the status of a vFiler (if the vfiler is running and if the network resources are configured)	stable	-
check_netapp_asup monitors the ASUP-log and alarms if failed transmissions or collections were found.	-	stable
check_netapp_license checks the filer for expiring (demo-)licenses.	stable	stable
check_netapp_nfs-persist checks for non-persistant NFS shares.	-	pre-alpha
check_netapp_process checks for runaway processes on a filer (as shown with the ps command).	-	stable
check_netapp_quotas monitors quotas on a NetApp-filer (cluster mode only).	-	stable
check_netapp_scrub sends an alarm if the last scrubs timestamp of an aggregate is over a certain age.	stable	stable
check_netapp_snapcenter checks the SnapCenter database for failed or missing jobs. This way alarms are sent immediately if backups do not run as expected.	-	stable
check_netapp_takeover sends an alarm if the storage failover facility is disabled or otherwise not active.	-	stable
check_netapp_time checks the filers NTP configuration (at least one ntp server must be configured) and measures the drift between the filers system-time and the monitoring server. Can alarm if that drift is getting too high.	-	stable
check_netapp_unused_lun checks for luns which are online but do not have an initiator connected.	-	beta
DiskCount counts the number of disks matching defineable criteria (disk-type, container (spare, ...), storage-pool). Mostly used to monitor the number of spare-disks of a certain type. `[help]`	-	stable	disk
DiskPathQuality hecks disk path qualities, reports i/o-error percentages and raises a CRITICAL error whenever an error percentage is above zero. `[help]`	alpha	beta	disk
DiskPaths checks if each disk has a given number and pattern of paths (A/B, B/A, ABAB, ABBA, ...). `[help]`	beta	stable	disk
FCPAdapter checks the operational status of all fcp adapters. `[help]`	-	stable	fcp-adapter
IfGrp checks if an interface-group has enough links in up-state to still be redundant. `[help]`	stable	stable	ifgroup (7m, ³) net-port (cm) ifconfig (7m, ³)
iSCSI monitors one or all iSCSI-adapters (online, offline, local, partner, error). `[help]`	unsupported	unsupported
Job checks for failed jobs. `[help]`	-	stable	job
LunAlignment searches for misaligned luns. Alarms if a certain number of misaligned luns is reached. `[help]`	-	stable	lun
LunSize checks the unused but allocated blocks inside of a LUN. Notfifys the admin if they exceed a certain number (he may than run an unmap procedure on vmware). `[help]`	stable	stable	lun
LunState checks the LUN-states. Alarms if they are offline or not mapped to an initiator. `[help]`	stable	stable	lun
NetInterface checks if a network interfaces current-port is not equal to its home-port (output of the CLI command `network interface show -is-home false`). Can also check it's operational mode (up/down). `[help]`	-	stable	net-interface (cm)
OvercommitAggr returns a list of aggregates together with their overcommitment in percent. Overcommitment is the relation between the aggregates size and the total of all its (thin provisioned) volumes sizes. `[help]`	stable	stable	volume aggregate
Raidstatus alarms if one of the RAIDs is degraded. `[help]`	stable	stable	aggregate
ReportIOPS reports how many iops are consumed by a given tenant. `[help]`	-	stable	volume
ReportSpace reports how much space in bytes are consumed by a given tenant. `[help]`	-	stable	volume
ServiceProcessor checks the status of the nodes service-processor and if they are correctly configured (autoupdate, IP-address). `[help]`	-	stable	service-processor
ShelfBay checks, the shelf- and disk-port status. Can alarm BYP-status disks. `[help]`	stable	stable	shelf-bay
Sis checks dedup-values (stale-fingerprint-percentage, run-time of last successfull operation). `[help]`	-	stable	sis
SisStatus find volumes whose compression or deduplication is not enabled. `[help]`	-	beta	sis
SnapMirrorMetrics checks and logs SnapMirrors (including type Vault): lag-time, last-transfer-duration, last-transfer-size `[help]`	-	stable	snap-mirror
SnapMirrorState checks and logs for SnapMirror (including type Vault): health, mirror-state `[help]`	-	stable	snap-mirror
SnapshotChangeRate calculates and monitors the change-rate (daily data change) of Snapshots in Gigabytes per day. `[help]`	-	pre-alpha	volume
SnapshotLessVolume searches for volumes which do not have snapshots. `[help]`	-	stable	volume
StorageUtilization Storage Utilization answers the question, “Am I effectively using the storage capacity available to my applications. `[help]`	stable	stable	aggregate
UnprotectedVolume checks for volumes not protected by SnapMirror. `[help]`	-	alpha	volume (³) snapmirror-destination (³)
UsageTrend checks the time how long ist would last until an aggregate or volume is full, if the trend of the last 48h (configurable) would continue. Checks both bytes and inodes. `[help]`	stable	stable	aggregate volume
VolumeAge searches for and flags volumes which have been created a (configurable) long time ago. An old age may be an indication for a forgotten and unused volume-clone. The logic can be also inverted to search for volumes with an exceptional short age (which have been created within the last day or so). `[help]`	alpha	alpha	volume
VolumeAutosize checks a volumes total-size and alerts when the volume is close to being full relative to the autosize maximum. `[help]`	stable	stable	volume
VolumeState checks the volume-states. Alarms if they are not online (configurable). `[help]`	stable	stable	volume
Vserver monitors the admin-state or the operational-status of a Vserver (running, stopped, inconsistent or defunct) `[help]`	-	stable	vserver

Base Bundle

Check	Status		depends on object(s)
Check	7m	cm	depends on object(s)
check_netapp7_head monitors the 7m-heads hardware objects (fans, NVRAM, power-supplies and the temperature-sensors)	stable	-
check_netapp_health monitors the system health. Sends an alarm if the system health status is anything other than 'ok'.	stable	stable
check_netapp_spare monitors the status of the spare-low condition (alarms if there is no suitable spare disk available).	unsupported	unsupported
Disk checks for failed, offline or unassigned disks on the filer. `[help]`	stable	stable	disk
DiskFailed checks for failed/offline disks on the filer.	unsupported	unsupported
Hardware checks the NVRAM, cooling-devices, temperature-sensors and power-supplies of the head and its shelves. Also checks all disks in the system, if they are in a broken, pre-failed or replacing state. `[help]`	unsupported	unsupported
Head monitors the heads hardware objects (fans, NVRAM, power-supplies, health-state, temperature-sensors) `[help]`	-	stable	head
NetPort checks if the network-interfaces are enabled or not `[help]`	-	stable	net-port
NetPort7m checks if the network-interfaces are enabled or not `[help]`	stable	-	ifconfig
Quotas sends an alarm, if one of the defined quotas are exceeded. One service-check can monitor the whole filer, since we use the internal definitions of the quota-system as thresholds. `[help]`	unsupported	unsupported
ShelfEnvironment checks, the shelf-status, power-supplys, temperature, fans, voltage-sensor and current-sensor on the shelves. `[help]`	stable	stable	shelf-environment
Snapshots checks, if the snap-reserve is still sufficient. Thresholds are set in percent; performance-data can be either in percent or absolute (Byte). Additional criteria are the age or name of the snapshot. This can be used for monitoring snapshot-backups and whether they are up to date or not. Also can be used to find snapshots related to a specific application like SNMV and check all volumes for left-over snapshots. `[help]`	stable	stable	vol_snapshot (³) aggr_snapshot (³) volume (³) aggregate (³) lun (³)
Uptime checks the seconds since last reboot. `[help]`	-	stable	head
Usage checks the used space in volumes and aggregates. Thresholds can be set in GB or percent. `[help]`	stable	stable	volume aggregate

MetroCluster Bundle

Check	Status		depends on object(s)
Check	7m	cm	depends on object(s)
check_netapp_mc_config checks a metro-clusters mode and configuration state.	-	stable
ClusterPeerHealth checks the health of cluster peer relationships by evaluating several ping- and health-status. `[help]`	-	stable	cluster-peer-health
MetroClusterVserver sends an alarm if the configuration state of a MetroCluster vserver changes to unhealthy. `[help]`	-	stable	metrocluster-vserver
SyncMirror checks the mirror-status on Metro Cluster aggregates. `[help]`	stable	stable	aggregate

Performance Bundle

Check	Status		depends on object(s)
Check	7m	cm	depends on object(s)
BadlyPerformingDisks checks all disks in a NetApp system or in a specific raid-group. If a certain number of them performes badly (=has a high utilization) an alarm is send. `[help]`	stable	stable	disk
BufferCache checks several metrics of the system buffer cache (=system memory) like Buffers being read, Buffers being written, Empty (unused) buffers, Buffers with modified data, Buffers associated with CP IO, ... `[help]`	stable	stable	bufcache
FlashCache checks several metrics of the external FlashCache (PAM II) like External cache hit rate, Average latency of read I/Os, Number of wafl buffers served off the external cache, ... `[help]`	stable	stable	ext_cache_obj
LunLatency Checks the 'latency' and 'operations per second' (ops) per LUN. Shows details for total, read, write and other. NetApp recommends monitoring latency as the primary performance indicator. `[help]`	stable	stable	lun
NVRAM checks data-rates and latency of the NVRAM. `[help]`	stable	stable	nvram
PerfAggr The area of application for this plugin is to monitor performance-relevant aspects of capacity management within an aggregate -in other words, the percentage of free blocks available for writes within an aggregate (ignoring all reservations, guarantees, or the like - just the free space available to the write allocator). This allows to proactively order additional storage for an aggregate before write performance is impacted. `[help]`	unsupported	unsupported
PerfAggregate checks the 'latency', 'transfer-rate' and other performance counters per aggregate. Shows details for total, read, write and other. Also averages and totals over all aggregates of the filer can be measured and monitored, which allows the monitoring of the aggregate-latency and aggretate-transfer-rate on the filer level. `[help]`	beta	stable	aggregate
PerfCpu checks one or all processors in a NetApp system for their utilization. `[help]`	stable	stable	processor
PerfDisk checks all disks in a NetApp system for their utilization (Percentage of time there was at least one outstanding request to the disk). Optional the check can be limited to the disks of a single aggregate. `[help]`	stable	stable	disk disk_constituent
PerfHostadapter checks and counts rates per host adapter (Fibre Channel, Serial Attached SCSI, and parallel SCSI). `[help]`	stable	stable	hostadapter
PerfIf checks and counts transfer-rates and errors per network-interface (ifnet). Especially useful for monitoring 10GbE-ports. `[help]`	stable	stable	ifnet
PerfLif checks and counts transfer-rates and errors per network-interface (lif) for DataONTAP 8.2.x. or higher. `[help]`	stable	stable	lif
PerfNic checks various performance counters of a NetApps physical network interface (NIC). Among them are crc/transmit-error-counters which can be used to detect errors on the physical network-layer. `[help]`	-	beta	nic_common
PerfQtree checks some ops-counters per q-tree (nfs-ops, cifs-ops, ...). `[help]`	alpha	stable	qtree
PerfSys checks various performance counters of the NetApp-system (mostly operations/second and transfer-rates). Counters supported: net_data_sent, dafs_ops, total_ops, disk_data_written, net_data_recv, cifs_ops, streaming_pkts, http_ops, nfs_ops, fcp_ops, disk_data_read, iscsi_ops `[help]`	stable	stable	system
PerfSysNode checks various performance counters of the NetApp-system (mostly operations/second and transfer-rates). Counters supported: net_data_sent, dafs_ops, total_ops, disk_data_written, net_data_recv, cifs_ops, streaming_pkts, http_ops, nfs_ops, fcp_ops, disk_data_read, iscsi_ops. The check evaluates these counters per Node and works only for DataONTAP 8.3 or later. `[help]`	beta	stable	system_node
PerfTcpIp checks CRC errors and packets send/received for both the IP and TCP layer. `[help]`	-	stable	tcp ip
PerfVolume checks the 'latency' and 'operations per second' (ops) per volume. Shows details for total, read, write and other. NetApp recommends monitoring latency as the primary performance indicator. `[help]`	stable	stable	volume
Wafl reads WAFL performance-counters like cp_count twice and calculates the rate of CPs per second. Different types of consistency-points (wafl-timer, back-to-back, ...) can be checked. The information gathered from this plugin corresponds to the CPty-column of 'sysstat -x 1'. `[help]`	stable	stable	wafl