process memory usage monitoring using glance

[Tim Killinger]

hello - help for a newbie SA learning performance monitoring....

When monitoring our system using glance, can someone explain the "RSS" column. Also the "RSS/VSS" column when you run show all the resources for a particular process?

Is this showing actual memory use by this process, and if so, what is the kernel parameter that sets the absolute limit for any particular process?

Thanks in advance!

[Pete Randall · Accepted solution]

Tim,

From the GUI version of Glance's online help:

RSS

The size (in KB) of resident memory for the process. This consists of text, data, stack, as well as the process' portion of shared memory regions (such as, shared libraries, text segments, and shared data).

Resident memory (RSS) is calculated as

RSS = sum of private region pages +

(sum of shared region pages / number of references)



Vss



This consists of the sum of the virtual set size of all private and shared memory regions used by this process. This metric is not affected by the reference count for those regions which are shared.

Note, a value of "na" may be shown for the swapper process.

This metric is specific to a process. If this metric is reported for a thread, the value for its associated process is given.


Pete

[Steven E. Protter]

A few issues to understand.

Memory from the HP-UX standpoint is memory plus swap.

The RSS figure is a figure that shows how much memory the process is using.

Processes don't really care how much memory your system has on it, they just take a chunk out of the total(ram plus swap). From a performance standpoint memory matters.

As far as an absolute limit for a process in the kernel, I don't think thats possible. If possible it would probably not be advisable. If a process requests memory, it needs it for some reason. If denied access a process is may crash or simply sleep and fail to get its work done.

I'm attaching some perf monitoring scripts that might provide some useful,non-glance data.

SEP