[geeks] operating systems to replace Solaris

[Jonathan Patschke]

On Mon, 11 Apr 2011, Dr Robert Pasken wrote:

> With the qualifier that I have a generator backup, which comes up to
> full load in 2 minutes and battery backup for 10 minutes. The worst
> Solaris uptime I ever have had is 10 months, linux is typically under 5
> days and BSD is 10's of days.

That sounds like hardware.  Windows gets better uptime than that on good
hardware.  Are you comparing Solaris next to BSD and Linux side-by-side on
the same machine?  Your experience just doesn't seem to relate to mine:


$ uname -srp && uptime
FreeBSD 6.2-RELEASE-p3 amd64
12:42PM  up 1456 days,  7:59, 1 user, load averages: 0.45, 0.12, 0.04

$ uname -srp && uptime
FreeBSD 7.0-RELEASE-p7 amd64
  1:42AM  up 114 days,  8:31, 2 users, load averages: 0.05, 0.03, 0.00

$ uname -srp && uptime
FreeBSD 7.1-RELEASE amd64
  1:42AM  up 248 days,  9:34, 2 users, load averages: 0.20, 0.52, 0.08

$ uname -srp && uptime
FreeBSD 8.0-RELEASE-p2 amd64
12:42PM  up 345 days, 20:55, 1 user, load averages: 1.07, 1.09, 1.08

$ uname -srm && uptime 
OpenBSD 4.7 amd64
12:42PM  up 223 days, 22:39, 1 user, load averages: 0.12, 0.09, 0.08

$ uname -srm && uptime
OpenBSD 4.2 i386
12:43PM  up 69 days, 7 mins, 1 user, load averages: 0.13, 0.14, 0.10


The two FreeBSD 7.x systems are installed at a semiconductor plant in
Taiwan.  I can assure you those systems get routine vibration, airflow,
and power abuse and frequently have their networks torn apart underneath
them.  I'd be nearly as comfortable running RHEL in place of FreeBSD
there, too.

To contrast, the Solaris boxes get jumpstarted about weekly because a
piece of software we run tickles a few bugs in STMF, causing it to slowly
eat itself and trash its configuration database.  The machines are
effectively appliances, so jumpstarting actually makes more sense than
recovering the configuration.

-- 
Jonathan Patschke |
Elgin, TX         | "He who is contented is rich." -- Lao Tzu
USA               |