Project: Samba Performance Monitoring

James Peach jorgar at gmail.com
Sun Jul 9 22:46:57 GMT 2006


On 7/10/06, jijjy81 at arcor.de <jijjy81 at arcor.de> wrote:
> hello,
>
> I'm a student of the University Duisburg-Essen and write my diploma
> about "predefined rules for samba performance monitoring".
>
> For one or two years two students of the University Duisburg-Essen
> and Materna in Dortmund started a project for the ZRWEST (a big service
> provider) in Düsseldorf.
>
> "Main goal of this project was to design and implement an
> integrated platform, which allows automated collection of specific
> performance information about the file serving solution and the overall
> performance of the underlying operating system, analysis of the collected
> performance data, based on predefined rules, and automated initiation of

You should look into PCP (http://oss.sgi.com/projects/pcp). PCP is a
metrics framework that lets you sample all sorts of performance and
status counters either locally or across a network. This lets you do things
like correlate the number of smbd processes with the total CPU usage across the
machine or per-cpu.

One part of PCP is a program calles pmie (performance metrics inference
engine). This has a small language that lets you trigger events when
based on the results of evaluating expressions consisting of PCP metrics.

> countermeasures in case of detected abnormalities. The main objective of
> the platform, further referred to as Integrated Samba Inspection Service
> (ISIS), was to be deployed in the network of a big public service provider
> and assist the network administrators in tracing and resolving problems
> with the file services."
>
> My task is now to develop predefined rules to detect abnormalities in
> real time.
>
> Firstly I've to understand the "normal" behaviour of a system like:
> %IDL (percentage of time the system spent in idlemode, averaged over all
> available cpu instances), system calls .... and  the behaviour of the
> processes (smbd) like: %CPU (percentage of recent CPU time used by the
> process), RSS, SIZE, IOC/s (input/output of characters per second)....!!!
>
> To get these values I try to repeating several times by downloading
> some different size of files from the samba server with the windows
> client. By collecting and analyzing with statistics methods, I found
> out that the values of each parameter (%CPU, IOC/s...) were not
> constant. After correspondence with Mr. Lendecke, the cause of this
> state is the windows client.
>
> Now my question: "why does the abnormalities of the values depends on
> the windows client??"

Client-side caching is the most likely cause.

-- 
James Peach | jorgar at gmail.com


More information about the samba-technical mailing list