snapdelete.alert is an alert script for mon that will automatically
delete the oldest snapshot off of a NetApp filer volume.  This is useful
in the situation where the snapshots are kept around for convenience and
regularly fill up enough to cause problems using the volume.  It requires
Theo Van Dinter's enhanced netappfree.monitor script, and rsh-like access to
the filer to execute the necessary snapshot commands. (rsh is available from
the filer by default ("options rsh.enable" and etc/hosts.equiv), but SSH may
be available if the filer has SecureAdmin enabled.  YMMV.)

Tested against OnTap 5.2.x, should work on later versions as well.

To install, simply place the alert script in the mon alerts directory.
The alert can be added to the mon config like any other alert, but it has a
few optional parameters that you may want to use:

	-M	This specifies that if a valid snapshot isn't available for
		deletion (they're all gone already, etc.), a mail should
		be sent out saying as much with a list of currently
		available snapshots.

	-m <mail address>
		This is a comma-seperated list of email addresses which
		should be contacted when a snapshot is deleted or (using
		-M) if no snapshot is eligible for deletion.

	-o <snap types>
		This is a comma-seperated list of snapshot types
		(weekly, nightly, hourly) which specifies an order
		of deletion.  ie: if "-o weekly,nightly,hourly" is
		specified, weekly snapshots are all removed first, then
		nightly snapshots, then hourly snapshots.  The default,
		if "-o" isn't specified, is to delete the oldest valid
		snapshot available.

	-r <rsh path/parameters>
		The default command for use against the NetApps is rsh (since
		it's available by default).  If you need to change the path
		or parameters ("/usr/bin/rsh -l root" by default), use "-r".
		For instance, on HPUX, the command is "remsh" instead of
		"rsh".

	-v <snap types>
		This is a comma-seperated list of snapshot types which are
		allowed to be removed.	For instance, if the configuration
		has "-v weekly,nightly", only weekly and nightly snapshots
		are considered for removal.  The default, if "-v" isn't
		specified is "weekly,nightly,hourly".

When configured, the mon configuration will look like:

watch toasters
	service snapfreespace
	description Check snapshot space on toasters
	interval 5m
	monitor netappfree.monitor -community=sysmon \
		-config=/usr/lib/mon/etc/netappsnapfree.cf
	period wd {Sun-Sat}
		alert snapdelete.alert -m it-alert@company.com

This will run netappfree.monitor every 5 minutes and check on the snapshot
space-only (per the netappsnapfree.cf file).  If any snapshot areas are
running low on space, snapdelete.alert is called and will mail
it-alert@company.com when a snapshot is deleted.  If there are no snapshots
to delete, we'll assume that another service (not shown) to check volume
space will alert if the snapshots get too large.

FYI: I haven't tested it, but the alert should handle a single
netappfree.monitor run for the volumes and snapshots.  The alert script will
ignore non-snapshot details from the monitor.