OpenBSD

[rant]
From time to time, people try to argue with me, and try and convince me that OpenBSD is a serious OS, one that I should waste my time administering. Okay, so it's got some credibility, sure. But does it have a good way to manage lots of machines? I'm not talking 5 or 6, I'm talking "Please expect a shipment of 74 servers next week" to add to a currently running datacenter. I'm talking paying for electricity by how many 20A circuits you use, and getting bandwidth 95th percentile.

Before you answer, the right answer is: Hell no.

I found more fuel to add to the "OpenBSD is the most unprofessional UNIX I've ever seen" department.

In my life, I tried OpenBSD. I really did. I fed it the best hardware I had. I gave it a decent shot.
It blew up in my face.

So, in the context of giving it a shot, I file a serious, well written bug report.


http://www.monkey.org/openbsd/archive/bugs/0112/msg00230.html

Number: 2284
Category: system
Synopsis: When doing heavy sustained disk activity, such as untarring bind9, the system kernel panics.
Confidential: no
Severity: critical
Priority: high
Responsible: bugs
State: open
Class: sw-bug
Submitter-Id: net
Arrival-Date: Fri Dec 28 20:00:01 MST 2001
Last-Modified:
Originator: Paul Timmins
Organization:
net
Release: 3.0
Environment:
Description:
cac0: Unable to alloc CCBsd0: Not Queued, Error 5
uvm_fault(0xe04f5670,0x1000,0,3) -> 1
Kernel Page Fault Trap, Code=0
Stopped at _pool_get+0x27c: movl %eax,0(%edx)
How-To-Repeat:
In my environment, untarring bind 9 (tar xzvf bind…tar.gz)
Fix:
None known to me.

Audit-Trail:
Unformatted:
environment: Freshly booted Compaq Proliant, Compaq SMART raid array, configured in RAID 5. Pentium II 400, 512k Cache, Running OpenBSD 3.0
System : OpenBSD 3.0
Architecture: OpenBSD.i386
Machine : i386


http://www.monkey.org/openbsd/archive/bugs/0112/msg00234.html

Synopsis: When doing heavy sustained disk activity, such as untarring bind9, the system kernel panics.

State-Changed-From-To: open-analyzed
State-Changed-By: mickey
State-Changed-When: Sat Dec 29 10:47:34 MST 2001
State-Changed-Why:
yes, i know it happens sometimes (;

Excuse me, it happens sometimes? That's a bug analysis? I'd expect that from something that's not supposed to be considered stable and useable, like perhaps a CVS HEAD of something, but this is the system kernel of a well known fork of BSD!

But it gets better….


2-3 years later
http://www.pantek.com/library/general/lists/openbsd.org/bugs/msg02357.html

Synopsis: When doing heavy sustained disk activity, such as untarring bind9, the system kernel panics.

State-Changed-From-To: analyzed->closed
State-Changed-By: deraadt
State-Changed-When: Fri Oct 3 18:18:03 MDT 2003
State-Changed-Why:
lots of changes have happened in that driver

Wait, my bug was closed WHY?!?!? BECAUSE LOTS OF CHANGES OCCURED IN A DRIVER.
Yep, that driver changed a lot. That obviously means the problem is fixed.

Please note that if you're gonna apologize for OpenBSD, you should read this first, to save me time.

(And yes, I know about pkg_add -r blah in FreeBSD, congratulations for finding one way to not suck TOO much.)

[/rant]

3 thoughts on “OpenBSD”

  1. I feel your pain.

    Some of this I know from first hand pain. Lots more I know due to the fact that I am in such close proximity to and I get to hear about the broken FreeBSD ports tree on a regular basis, mostly due to the fact that this fine institution has been running it for years and half our time dealing with it has been with band-aids.

    Also I actually bothered to read through your entire thread from back in 2003, and found it very amusing that the guy at the bottom was bragging about an OpenBSD system with 3 weeks of up time.

Leave a Reply