Too much uptime !

[Rabellino Sergio]

    [ The following text is in the "ISO-8859-1" character set. ]
    [ Your display is set for the "US-ASCII" character set.  ]
    [ Some characters may be displayed incorrectly. ]

Julian Field wrote:

>-----BEGIN PGP SIGNED MESSAGE-----
>Hash: SHA1
>
>High load average figures from "uptime" are not an indication at all 
>that your system is overloaded. It gives the average number of processes 
>in the run queue. If a process is waiting for disk, it adds 1. If a 
>process is waiting for network response, it adds 1. It doesn't mean 
>anything is overloaded at all.
>
>  
>
Yes, I know, but why there are 2 MS processes always in run-queue ?
They don't go to sleep  for a while between the queue scanning ?

>A load average of 3 means there are an average of 3 processes waiting 
>for cpu, disk or network. Since you have 4 CPUs then 1 of them is idle. 
>A load of 12 to 15 on a hard-working single cpu system is perfectly 
>common, it just means that all your system resources are being 
>thoroughly used.
>
>I usually recommend 5 children per CPU. So you should start by setting 
>it to 20.
>
>  
>
I was used to see a load-average between 0->1 so now I noticed a change
in this parameter and I need to understand why it happen.

>The Queue Scan Interval won't make any difference, I might remove it 
>some time. It only affects the behaviour of the system when there are 
>very few incoming messages. In all other situations it will process as 
>many messages as it can without waiting for anything. It doesn't mean 
>that MailScanner always waits 40 seconds to collect messages. If there 
>are any messages waiting at all, then they are immediately processed. 
>Leave it at 6.
>
>  
>
ok.

>With a load average of 3 on a box with 4 cpu's, you've wasted your money 
>as at least 1 cpu isn't even doing anything. A load average of 30 or 40 
>would mean your machine is working for its dinner.
>
>
>  
>
got it, captain!

>>Dear list,
>>i'm getting an high uptime on my mailscanner box, and this is a new 
>>behaviour for me.
>>
>>The load average persist over 3.00 and my mail traffic is not so high.
>>My actual config is
>>
>>MailScanner Rel4.45.4 - McAfee VirusScan - ClamAV cmd line - SA 3.0.4
>>Max Children = 5
>>Queue Scan Interval = 40
>>
>>I've changed The Scan Interval from 5 seconds to 40 and the max 
>>children from 10 to 5, but without any significant change in the load.
>>My Solaris 9 box is an V440 with 4GB of ram and 4 ultrasparc CPU.
>>
>>Where I need to investigate ? From a simple check, seems that 2 
>>MailScanner processes are statically  in O-state and three are in 
>>either R/S state: it's normal ?
>>
>>    
>>
>
>- -- 
>Julian Field
>www.MailScanner.info
>Buy the MailScanner book at www.MailScanner.info/store
>Professional Support Services at www.MailScanner.biz
>MailScanner thanks transtec Computers for their support
>
>PGP footprint: EE81 D763 3DB0 0BFD E1DC 7222 11F6 5947 1415 B654
>
>
>-----BEGIN PGP SIGNATURE-----
>Version: PGP Desktop 9.0.2 (Build 2424)
>
>iQA/AwUBQ0qtYBH2WUcUFbZUEQLMEACgiIawJOB95M8H1O+Ik7RrzcJRf88AoOvF
>RuE7jUyv6vT4UU7ax1Tb08rI
>=AYAX
>-----END PGP SIGNATURE-----
>
>  
>


-- 
Dott. Mag. Sergio Rabellino 

 Technical Staff
 Department of Computer Science
 University of Torino (Italy)

http://www.di.unito.it/~rabser
Tel. +39-0116706701
Fax. +39-011751603

------------------------ MailScanner list ------------------------
To unsubscribe, email jiscmail at jiscmail.ac.uk with the words:
'leave mailscanner' in the body of the email.
Before posting, read the Wiki (http://wiki.mailscanner.info/) and
the archives (http://www.jiscmail.ac.uk/lists/mailscanner.html).

Support MailScanner development - buy the book off the website!