The Unix Heritage Society mailing list
 help / color / mirror / Atom feed
From: Tony Finch <dot@dotat.at>
To: Dan Stromberg <drsalists@gmail.com>
Cc: The Eunuchs Hysterical Society <tuhs@tuhs.org>
Subject: Re: [TUHS] Zombified SCO comes back from the dead, brings trial back to life against IBM
Date: Sun, 4 Apr 2021 17:18:16 +0100	[thread overview]
Message-ID: <3c7d43cd-3f3d-bbcc-ea93-f27289935d5c@dotat.at> (raw)
In-Reply-To: <CAGGBd_r8JzjhUn-QjfJ2d_Vy_VLzb6jE1k2MPPdjRYS1oqJL4A@mail.gmail.com>

Dan Stromberg <drsalists@gmail.com> wrote:
>
> Please see https://stromberg.dnsalias.org/~strombrg/why-solaris.html for a
> list of benefits of 5.5 over 4.1.x.

A few of those are distinctly mixed blessings :-)

I learned about Solaris's in-kernel telnet the hard way, trying to debug a
system on which telnet would not work properly after rebooting. I think I
worked it out by trussing telnetd, which showed me that it was trying to
load a kernel module, and failing because we had chrooted it. A bit of
searching around told me what the module was for.

The Solaris automounter was neat - we did clever things with the userland
part, and it mostly performed well because the fast path was in the
kernel. Except that it didn't have a cache for ENOENT directory entries,
so requests for nonexistent paths always took a trip to userland, and (as
far as I could tell) these requests were serialized, and if you managed to
overload the automounter it would start returning bogus errors - I think
it was EPERM. Our Apache httpd started randomly returning 403 errors when
it tried to access a missing .htaccess (which it did on every request!)
and got EPERM instead.

The name service cache was also problematic. It was probably useful for
getpwnam() etc. on systems using NIS, but it behaved very badly on systems
that sent a lot of DNS queries through nscd. It worked much better for us
if we simply prevented it from running at all. And I'm not entirely sure
it was all that great for NIS: another problem I heard of second hand was
password files getting truncated when lots of users were trying to change
their passwords at the same time. My colleagues fixed that by ensuring all
password changes happened on the NIS master with suitable locking, and the
password files were periodically replicated with rdist instead of using
more normal NIS machinery.

Tony.
-- 
f.anthony.n.finch  <dot@dotat.at>  https://dotat.at/
Shetland Isles: West 7 or gale 8, veering northwest gale 8 or severe
gale 9, occasionally storm 10 later. Moderate or rough, becoming very
rough in sheltered east, otherwise very rough becoming high or very
high. Rain then squally snow showers. Moderate or poor, occasionally
good.


  reply	other threads:[~2021-04-04 16:18 UTC|newest]

Thread overview: 86+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-04-01 14:50 Josh Good
2021-04-01 15:12 ` Warner Losh
2021-04-01 15:27   ` Josh Good
2021-04-01 15:33     ` Larry McVoy
2021-04-01 16:14       ` Kevin Bowling
2021-04-01 16:26         ` John Cowan
2021-04-01 17:54       ` Thomas Paulsen
2021-04-01 16:27     ` Nemo Nusquam
2021-04-02  2:16     ` Kevin Bowling
2021-04-02  3:52   ` Wesley Parish
2021-04-02  5:26     ` Kevin Bowling
2021-04-02 16:03     ` Clem Cole
2021-04-02 16:11       ` Larry McVoy
2021-04-02 16:39       ` Heinz Lycklama
2021-04-02 17:14         ` Clem Cole
2021-04-02 17:17       ` [TUHS] AIX repeat [was " Charles H Sauer
2021-04-03  1:24       ` [TUHS] " Wesley Parish
2021-04-04  2:46     ` Adam Thornton
2021-04-04  2:50       ` Adam Thornton
2021-04-04  5:29         ` [TUHS] How to Kill a Technical Conference (was: Zombified SCO comes back from the dead, brings trial back to life against IBM) G. Branden Robinson
2021-04-04 18:22           ` Clem Cole
2021-04-04 20:54             ` Richard Salz
2021-04-04 21:11             ` Clem Cole
2021-04-05  0:36             ` John Cowan
2021-04-05  2:19               ` Warner Losh
2021-04-05 18:07                 ` [TUHS] How to Kill a Technical Conference John Gilmore
2021-04-05 19:30                   ` Clem Cole
2021-04-05 20:34                     ` Richard Salz
2021-04-05 20:42                       ` William Cheswick
2021-04-06  4:37                       ` Ed Bradford
2021-04-05 20:39                     ` Larry McVoy
2021-04-05 21:11                       ` Theodore Ts'o
2021-04-05 21:17                         ` Dan Cross
2021-04-06 15:39                     ` Theodore Ts'o
2021-04-06  5:49                   ` Dave Horsfall
2021-04-05  7:48               ` [TUHS] Whither Usenix [was How To Kill A Technical Conference] arnold
2021-04-05 14:05             ` [TUHS] How to Kill a Technical Conference (was: Zombified SCO comes back from the dead, brings trial back to life against IBM) Theodore Ts'o
2021-04-05 22:26               ` David Arnold
2021-04-04 23:30           ` A. P. Garcia
2021-04-04  3:41       ` [TUHS] Zombified SCO comes back from the dead, brings trial back to life against IBM Gregg Levine
2021-04-04  3:57         ` Adam Thornton
2021-04-02  5:41 ` David Arnold
2021-04-02  6:09   ` Steve Nickolas
2021-04-02  7:00     ` arnold
2021-04-02  9:53       ` Steve Nickolas
2021-04-02 10:26         ` arnold
2021-04-02 14:02           ` Josh Good
2021-04-02 14:17             ` Steve Nickolas
2021-04-02 15:16               ` Larry McVoy
2021-04-02 15:28                 ` Fabio Scotoni
2021-04-03  1:50                 ` Dave Horsfall
2021-04-03  1:55                   ` Warner Losh
2021-04-03  2:23                   ` Larry McVoy
2021-04-03  2:34                     ` Earl Baugh
2021-04-03  6:16                   ` Dan Stromberg
2021-04-04 16:18                     ` Tony Finch [this message]
2021-04-04  1:48                   ` David Arnold
2021-04-04  2:23                     ` Larry McVoy
2021-04-04  8:55                       ` Josh Good
2021-04-04 14:43                         ` Michael Parson
2021-04-04 15:36                         ` Warner Losh
2021-04-04 16:15                           ` Clem Cole
2021-04-04 22:25                             ` David Arnold
2021-04-04 22:55                               ` Clem Cole
2021-04-05  2:30                                 ` Dave Horsfall
2021-04-04 23:00                               ` Bakul Shah
2021-04-04 23:33                                 ` Clem Cole
2021-04-05  1:34                                   ` Bakul Shah
2021-04-05  2:58                                     ` Kenneth Goodwin
2021-04-05 12:35                                       ` John Cowan
2021-04-05 20:44                                   ` Kevin Bowling
2021-04-04 23:34                               ` Josh Good
2021-04-04 20:08                         ` Lyndon Nerenberg (VE7TFX/VE6BBM)
2021-04-04 21:00                           ` Jon Steinhart
2021-04-04 21:40                             ` Clem Cole
2021-04-04 21:54                               ` Lyndon Nerenberg (VE7TFX/VE6BBM)
2021-04-04 22:02                                 ` Jon Steinhart
2021-04-04 21:58                               ` Clem Cole
2021-04-04 23:48                             ` Dave Horsfall
2021-04-04 23:53                               ` Larry McVoy
2021-04-07  5:15                             ` Dan Stromberg
2021-04-05 13:37                       ` Theodore Ts'o
2021-04-07  1:52                         ` Adam Thornton
2021-04-02 15:25               ` Josh Good
2021-04-03  3:10               ` John Cowan
2021-04-02 16:40 ` Boyd Lynn Gerber

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=3c7d43cd-3f3d-bbcc-ea93-f27289935d5c@dotat.at \
    --to=dot@dotat.at \
    --cc=drsalists@gmail.com \
    --cc=tuhs@tuhs.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).