You're seeing load avgs that high, eh? And I assume to individual
vservers are still performing well?
I don't think I've ever seen a machine that wasn't nearly deadlocked
with load averages anywhere near that high...So I'm guessing this is a
sign of the flaws in load avg calculation.
If you don't mind my asking, what does a 'vmstat 1' say about your CPU
user/system usage and CPU idle during a high load period? I think that
tends to be a better display of what load the box is seeing and how much
more it can handle.
Kevin Druet wrote:
> we do use vsd commercially, we are using version 1.4.6 and dont have
> support for quite a few options. ( Mod BIND, SSL, MySQL)
> We are also running it on Debian.
> Here is the info for 1 Dell Poweredge 4400 Dual PIII 600 Server 512MB
> ECC, 4 Ultra 160 10K RPM Drives, hosting over 400 domains, and 20K +
> users.
>
> Here is a copy and paste of an uptime,
>
> [root@calinux2 /root]# uptime
> 3:22pm up 126 days, 8:53, 1 user, load average: 47.03, 29.07, 53.11
> [root@calinux2 /root]#
>
> This server has 3 NICS installed with 212 aliased IP addresses on it.
>
> what type of hardware are the 966's running in ? is it server
> architecture, or workstation architecture ?
>
> We speculate that if we increase or processors to dual gigahertz that we
> will be able to support MySQL back ends on the same server with no other
> upgrades.
>
>
> Regards,
> Kevin Druet
--
Joe Cooper <[EMAIL PROTECTED]>
Affordable Web Caching Proxy Appliances
http://www.swelltech.com
------------------------- The freeVSD Support List --------------------------
Subscribe: mailto:[EMAIL PROTECTED]?body=subscribe%20freevsd-support
Unsubscribe: mailto:[EMAIL PROTECTED]?body=unsubscribe%20freevsd-support
Archives: http://freevsd.org/support/mail-archives/freevsd-support
-----------------------------------------------------------------------------
