[prev in list] [next in list] [prev in thread] [next in thread] 

List:       freebsd-smp
Subject:    Re: problem with smp in 4.1-STABLE ... or just 4.1-stable itself
From:       The Hermit Hacker <scrappy () hub ! org>
Date:       2000-08-17 3:02:05
[Download RAW message or body]


later this aft, I found the problem ... it seems that running webalizer
against the logs of one of my domains genrated a 'dnscache' file that is
>1gig in size.  When webalizer runs, it loads the whole damn thing up into
memory ... before things locked up one of these times, I got to the point
where I could see that swap was something like 800+Meg used and rising
*roll eyes*



On Wed, 16 Aug 2000, The Hermit Hacker wrote:

> 
> wow, just hung my machine up solid twice in about a half hour period ...
> 
> I just upgraded my machine to a Dual-PIII 700 on an ASUS P2B-D motherboard
> ... everything has been running great the past 2 days, things just fly
> along, hardly ever see the load average rise about 1 ...
> 
> today, I'm working on improving stats generation for my virtual hosts and
> wrote a simple script that goes through and runs webalizer against any
> domain that has had activity since the last run ...
> 
> basically it checks a domain, runs webalizer in the background and goes
> off to the next one ...
> 
> as soon as I start the script, the load avg climbs relatively quickly
> (last one I saw was 26 and rising), the machine gets more and more lag'd
> until it just locks up solid ...
> 
> I'm running remotely, and have the serial console configured, but altho I
> can see the machine boot, get to single user mode and all that, as soon as
> I goes multi-user my keyboard no longer works, so I can't get to DDB ...
> 
> I'm running 4.1-STABLE code as of Monday night ... 
> 
> I hate to do it, but I'm going to try and kill it once more with top
> running and see if anything odd comes up in there ... yup, third time is
> the charm ... all when I do my stats run.  top shows:
> 
> last pid: 13913;  load averages: 71.25, 37.03, 15.564                    up 0+00:09:43  14:52:23
> 262 processes: 2 running, 258 sleeping, 2 zombie
> CPU states: 37.1% user,  0.0% nice, 45.3% system,  1.3% interrupt, 16.3% idle
> Mem: 368M Active, 42M Inact, 53M Wired, 2244K Cache, 61M Buf, 37M Free
> Swap: 2047M Total, 602M Used, 1446M Free, 29% Inuse, 12M In
> 
>   PID USERNAME   PRI NICE  SIZE    RES STATE  C   TIME   WCPU    CPU COMMAND
> 13513 corey        2   0  2220K  1324K select 1   0:03 32.17% 11.67% top
> 13512 jeff         2   0  2220K  1316K select 0   0:03 31.77% 11.52% top
>   918 scrappy     64   0  2208K   704K CPU1   1   0:11 26.25%  9.52% top
>  8568 root       -22   0  1026M   218M swread 0   0:16  4.40%  4.39% webalizer
>   610 root         2   0  2916K  1128K select 1   0:12  2.15%  0.78% named
> 13535 nobody      -6   0  3724K  3128K pipdwt 0   0:00  3.01%  0.78% perl
> 13534 nobody      -6   0  3724K  3128K pipdwt 1   0:00  2.82%  0.73% perl
>  1485 nobody       2   0  8956K   904K sbwait 0   0:01  1.75%  0.63% libhttpsd.
> 14003 root        64   0   864K   508K CPU0   0   0:00  0.00%  0.00% sh
> 
> I have 768Meg of RAM on this machine, and >2gig of swap space ... I
> actually saw a loadavg of 90 while watching, but couldn't capture it ...
> 
> does 4.1-STABLE have any known problems with not being able to handle
> "bursts" of processes? *raised eyebrow*  I'm suspecting the other
> lock(s) were just high load too, now that I watch top ... 
> 
> Comments?
> 
> Marc G. Fournier                   ICQ#7615664               IRC Nick: Scrappy
> Systems Administrator @ hub.org 
> primary: scrappy@hub.org           secondary: scrappy@{freebsd|postgresql}.org 
> 
> 
> 
> To Unsubscribe: send mail to majordomo@FreeBSD.org
> with "unsubscribe freebsd-stable" in the body of the message
> 

Marc G. Fournier                   ICQ#7615664               IRC Nick: Scrappy
Systems Administrator @ hub.org 
primary: scrappy@hub.org           secondary: scrappy@{freebsd|postgresql}.org 



To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-smp" in the body of the message

[prev in list] [next in list] [prev in thread] [next in thread]