Re: Hard-limits for kern.maxproc
Re: Hard-limits for kern.maxproc
- Subject: Re: Hard-limits for kern.maxproc
- From: Amanda Walker <email@hidden>
- Date: Fri, 6 Feb 2009 10:04:24 -0800
On Feb 6, 2009, at 7:55 AM, Rick Macklem wrote:
Although I suspect much of what you say is accurate, I will note that
Mr. Jobs went on and on during one of his keynotes (on their web site)
about Xserve (comparing it to a stack of Dells in a rack) and showing
a picture of a something nicknamed the "aquarium", which was a
relatively
large cluster of Xserves. (The aquarium looked like a "serious
production
environment" to me.)
I will admit to being biased towards the large end of the scale--I
think of "serious production environment" as "lots of servers across
multiple data centers". I am sure that many people use clusters of
XServes for serious work--that wasn't the sense I meant. I was
thinking more along the lines of "customers are paying you money to
meet an SLA, where every minute of downtime, every reboot, and every
time you have to send a tech out with a crash cart to sit in front of
the machine to do something, it costs you real money."
I will say that the Intel XServes are a nice advance over the XServe
G5 in this regard. Adding LOM support helped a lot, even if the IPMI
implementation is a bit quirky.
--Amanda
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Darwin-kernel mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden