sar -v

[Unix Administrator_6]

what is it? I cant find it in my man pages, but it gives me this output:

00:00:00 text-sz ov proc-sz ov inod-sz ov file-sz ov
13:20:00 N/A N/A 420/2088 0 2680/2680 0 2143/8010 0
13:40:00 N/A N/A 350/2088 0 2680/2680 0 1828/8010 0
14:00:00 N/A N/A 388/2088 0 2680/2680 0 1966/8010 0
14:00:01 N/A N/A 398/2088 0 2679/2680 0 1996/8010 0

Does the middle column (2679/2680) mean my inodes are maxed out or close to it? I dont have any errors in my syslog or anywhere else if thats the case.
I just wasnt sue what I was looking at and someone else is saying that its my inodes, and I wanted to be sure.

[Brian M. Fisher · Accepted solution]

It is not uncommon to see what appears to be a ?maxed out? value for
inode-sz column. The number to the left of the / is the number of
inodes open in the inode table cache and the number on the right is the maximum
number of inodes that can be open in the inode table cache determined from the
value of ninode in the running kernel. The HP-UX OS actually tries to
keep this value at the maximum for performance reasons. As more inodes are
cached, the inode retrievals (on average) will be faster. Seeing this value in
inode-sz to be equal to your ninode value is not something to be overly
concerned about. The system will maintain the cache and add/delete inode
entries as needed. This is unlike the proc-sz and file-sz columns
from the sar output which show hard limits. When these limits are
reached, new processes cannot not be started or additional files cannot be
opened. The inode-sz column refers to a cached table, and it is expected
that having this value ?maxed out? should not prevent users on the system from
extracting inode information from inodes not available in the cache. That
being said, tuning ninode to be a smaller or larger value to allow for a
smaller or larger inode cache table can have a neglibile effect on performance
in some environments.

Per document#KBRC00001929
http://us-support2.external.hp.com/cki/bin/doc.pl/

Brian
<*(((>< er

[Rick Garland]

I find it the man page.
List status of text, process, inode, and file tables

[Bill Hassell]

Just a clarification about the indoe table. It will always have an entry for every file currently open on the system (no matter how many times the file has been opened by multiple processes), and 'remembers' old inode information in hopes that a process will soon need to open this recently used file. The kernel parameter for this is ninode. The parameter nfile refers to all open file handles, one for file open even if it is already open by another process.

Unfortuantely, sar (and Glance) cannot obtain information on how many entries are actually in use versus older entries that can be resused, so the number is always high, typically 99-100% of the max. If the table ever does fill up with unique open files hoever, the error "inode: table is full" will be displayed on the console.

Since you can't really determine how full the inode table is, you have to take a guess. For instance, if ninode = nfile, then you won't experience any overflows. However, the inode table is only used for HFS filesystems, not VXFS which uses a different mechanism.

Since HFS is not the default, the majority of your filesystems may be VXFS and only /stand needs to use this cache, so setting ninode to 1000 would be more than adequate.

[Jim Welch]

I use the overflow flags in the sar output to see if I'm really running out of entries vs just having unused entries cached.