The output of sar -v on one of my servers shows inode-sz 0/5168
The server is at 11.0 and the filsystems are set up vxfs. I don't understand how current number of inode cache can be 0. This server is also running 4 oracle databases.
Not to worry, ninode refers to hfs filesystems and I'll bet the only hfs filesystem on your box is /stand. You could probably reduce ninode to 200 or so and be fine (and save a few resources as well).
It does not apply to JFS, that's why you see 0. You would start to worry if it's registering high usage value if the only FS you have that is non-JFS is the boot LV (/stand).
The reason I am questioning this is because all of our HPUX servers, over 100 are set up this way (yes, only /stand as jfs) and this server is the only one that reports 0. It makes me wonder if something is not set up properly.
From the book .. The "inod-sz" represent the # of of ACTIVE inodes (including HFS & JFS inodes). The 1st value (before /) is the current size of the inode cache and the 2nd value (after /) is the maximum size of the inode cache. This maximum size reflects the maximum size of the HFS inode cache (from kernel value "ninode"). ... You got me !! You should see some "active" inodes registering in the sar output at least. What does glance shows ?
I'm don't see this on my 11.00 systems at all but a few of my 11i servers does show 0. On my 11.00 server .. # what /usr/sbin/sar gives $Revision: 82.1.1.12 $ PATCH_11_00: sar.o 00/12/01
whereas on my 11i server .. # what /usr/sbin/sar gives $Revision: B.11.11_LR Fri Oct 27 00:31:45 PDT 2000 $
I'm thinking "sar" might need to be patched. You may want to try keyword search "sar" in the patch DB to see if anything come up at all.
