Re: i5 Youngsters

On 1/15/08, Jones, John (US) <John.Jones@xxxxxxxxxx> wrote:

I'll agree that all servers have the _potential_ to go down for
unplanned reasons in addition to planned reasons. But I'd still argue
that a System i will have fewer unplanned outages than a Windows server.

Why?

Hardware quality is, assuming you purchase a decent Windows server, the same.

Software quality is the topic right now. Our internal Systems have
been running without unplanned downtime since i'm working here (3
years). That's one production System i (270 for long, now a 515) and
three production System x (330, 346, 3650) running Windows Server 2003
R2, and two production System x (306m, 3250) running Debian Linux.

All of them had downtime due to software upgrades or hardware
upgrades. Not a single one crashed in these three years. None of them
require attention - i get mails from all of them about backup
completion, and they run fine when unattended (which happens when i'm
overloaded with more important work - which is currently not the case
;).

load balance. To achieve similar uptime we've added 4 new vendors to
the mix, 2 new operating systems, a front-end processor, an additional
rack in the data center, more networking gear (new firewalls & more
switch capacity), a backup appliance, and so on. And we'll have to
replicate this in our DR center eventually.

No. This is a "better" setup than a single System i running
everything. First, you have hardware redundancy - a server fails? No
problem, just swap it out. If your System i fails, you'll have to
switch to your DR site.

You need more processing power? Just add a few front end servers. With
the System i, you'd need to buy a new CEC and replace the current one
(which would probably never happen, far too expensive).