Infrastructure Monitoring

Overview:
I'm looking for a way to get usage data for all mount points (UFS,VxFS & NFS) on a Solaris system. I'm starting with `df` but would like to find something that might be more light-weight.

Background:
Using SNMP is not an option in this case, instead I need to use basic built-in commands to get the data. The commands must be present on SPARC and x86 platforms and our minimum release level is Solaris 8, newest is 10. I will be writing a script or program (don't need help with that) to get the disk usage statistics and then report on partitions that are low on space. The script will run automatically and on systems where mount points tend to hang or disappear without notice so I'll need to fork df for each mount. The additional processes, some which may get orphaned, might be a resource problem so I want to at least minimize the resource footprint of each command that is run.

I am fully ready to accept that`df` may already be the best way to approach this but I wanted to ping the community just in case.

In you write in C or perl you can call statvfs() directly. For a shell script, df is the what I would use.

That is a good suggestion, thank you. I'm going to try to not write new software at this point but if things get ugly I may not have a choice.