[rescue] Spare Sun box for cheap/free?
mark at hermsdorfer.net
Fri Dec 14 12:20:10 CST 2007
On Dec 14, 2007 12:31 PM, Brooke Gravitt <gravitt at gmail.com> wrote:
> On Dec 14, 2007 12:19 PM, Mark Hermsdorfer <mark at hermsdorfer.net> wrote:
> > Websphere should run fine on 300Mhz procs... it's the ram you want to
> > out for. I would not want to put WAS on a server with less than 1GB of
> > you might get it going with less but I bet it would be painful... even
> for a
> > play environment.
> > The nodeagent JVM and deployment manager JVM both run 256MB max heap by
> > default. Then at least 1 application server JVM is another 128-256MB of
> > heap depending on how you configure it. Of course the JVMs likely won't
> > out the heap... particularly on the dmgr and node agent.
> Tell me about it. That's one of our team's duties, managing IHS/WAS
> for the Enterprise. We at least made top 5 for uptime over the
> post-thanksgiving sale week. We beat the Apple Store! ( I'm at Macy's
> Systems - Macys.com & Bloomindales.com )
> rescue list - http://www.sunhelp.org/mailman/listinfo/rescue
Congrats on the uptime.
Sounds like we have somewhat similar jobs... I do operations work on IHS/WAS
for a fortune 500 pharma company. The applications in the DMZ are fairly
simple apps often just a webcontainer a few servlets and no ejbs. I'm sure
we also have much lower user loads than a site like Macys. Some of our more
'fun' applications are internal apps used for financial analysis.
This has been my first real job out of university, been here for almost 2
years. I have to say am very impressed with the stability of
Solaris/IHS/WAS it seems like a fine platform to me.
The things that cause us the most pain are normally related to other teams:
oracle, the MQ group, developers using poor coding techniques then blaming
operations when the app fails under load. Ahh.. you have to love operations
More information about the rescue