Date: Thu, 2 Apr 1998 18:30:32 +0100 From: nik@iii.co.uk To: Sue Blake <sue@welearn.com.au> Subject: Re: login_getclass unknown class Message-ID: <19980402183032.37825@iii.co.uk> In-Reply-To: <19980403023433.63772@welearn.com.au>; from Sue Blake on Fri, Apr 03, 1998 at 02:34:33AM %2B1000 References: <Pine.BSF.3.91.980401173352.19507A-100000@user.xtdl.com> <199804020125.TAA08568@darkstar.connect.com> <19980402113719.25471@welearn.com.au> <19980402095351.50685@iii.co.uk> <19980403023433.63772@welearn.com.au>
index | next in thread | previous in thread | raw e-mail
On Fri, Apr 03, 1998 at 02:34:33AM +1000, Sue Blake wrote:
> > On Thu, Apr 02, 1998 at 11:37:19AM +1000, Sue Blake wrote:
> > > On Wed, Apr 01, 1998 at 07:25:46PM -0600, Frank Pawlak wrote:
> > > > You might want to check out the errata for release 2.2.2
> > >
> > > That won't necessarily help.
> > >
> > > Why doesn't somebody just tell this guy where the file he is missing
> > > can be found, and leave it at that.
> >
> > I think this is part of the "Give a man a fish, you feed him for a day.
> > Teach a man to fish, you feed him for a lifetime." philosophy, although
> > (IMHO) Frank could have phrased his answer a little better.
>
> I wholeheartedly support that philosophy, but when properly applied it is
> easily distinguishable from the punitive teasing responses we sometimes see.
I'm not sure that Frank's response comes under that heading. He could
probably have been a little more lucid, but that's certainly enough for
the enterprising reader to go the search engine on the site and type
in "errata 2.2.2" and see what comes back.
That said, it would help if there was a direct link to the release
information on the first page of the site. As I've said in other e-mail,
I think the site (certainly the first page) is not aimed sufficiently well
at people who want to find information. Of course, I haven't had the time
to do anything else about it :-(
> If the helpful person has clearly not bothered to do the act that they are
> demanding of the other, they do not earn my respect. And if that omission
> leads to incorrect advice being given and/or gross and unproductive
> inconvenience to the person who is already having a hard time, then I start
> to lose my cool.
Uh, now I'm confused. As far as I can tell, Frank did check out the
errata to see if it solved the problem. His 'mistake' was that he checked
the errata file on the web site, which doesn't indicate that which changes
have been added since the release went live. I've just had a look at
the others on the web site, and they don't either. This probably needs
mentioning to Doug, so he can include it for the next release.
> > I just went to the search engine, and put "Login_getclass" in the mailing
> > list search box.
>
> OK, so you know what to do and how to squeeeeeeeeeze the best out of the
> search.
:-) We have different ideas about what constitutes a difficult search,
but that's OK.
> And you've been considerate enough to show us how you'd go about it,
> thanks. It's a lot more than most bother to offer. I think we need to work
> through a lot of examples to become good at searching.
That's cool. And would probably be (another) good thing for the folks
in -newbies to think about. The more examples the better, and someone
to take the time to write them up. All it takes is volunteers.
> Now I know, if someone wants my advice on "hey, I did everything right but
> it's giving this error, how do I fix it" I am to give them no hints but just
> a huuuuge list of little tasks that will do them ever so much good, and
> teach them some useful skills in the process, like reading the handbook,
> faq, searching the mail archives _and_ the web pages, reading all of the
> .TXT files on the FTP site, looking for man pages and other documents,
> tutorials, books,... Sure, it's a lot of work but by doing it they will
> learn what they should have known to stop them making the same mistake
> again, right? Wrong. In this case it was nobody's mistake and nobody should
> have had to be punished or, ahem, retrained.
By that argument there's no point in maintaining the handbook, FAQ or
mailing list archives, since no one's going to use them.
This is not rocket science, but it's not "Press a, then b, then c" either.
<snip, because. . .>
> On the other hand, some newbies are the most lazy inconsiderate greedy twits
> that you could possibly imagine, and they feel quite happy about asking
> numerous simple questions each day to save them the trouble of looking
> things up or attempting think for themselves. And some are quite blatant
> about it. When these people are pandered to, as I've watched happening in at
> least one case recently, I get just as annoyed.
I think you might just have hit the nail on the head. As I think I mentioned
earlier, if someone's gone to the trouble of looking things up, and either
can't find what they're looking for, or don't understand what they've
found, and I can help, then I'll try and put together a longish answer.
There was one of those recently about permissions for a web server. The
chap knew where to look (chmod, chown and so on) and understood how they
worked, but didn't have the experience to put it all together.
If, on the other hand, he'd just posted "How do I set up my webserver
properly?" I'd probably have ignored it.
I'll just quote the original message that sparked all this off:
> What and why am I getting
> Login_getclass unknown class "root"
>
> I have cvsup-ed the latest stable and have done make world.
> I though 2.2.5 would correct this problem. I have been running
> 2.2.2 . I havn't rebuilt my kernel yet.
Straight off, this tells me that this person hasn't properly remade the
world or read the documentation on how to do it in the Handbook. Because
if they had, they would have merged in the changes to /etc (including
login.conf) as a part of the process, and they wouldn't be seeing this
problem anyway.
> > For jollies, I did the same search on the web pages.
>
> You did a search on the web pages?? Whatever would possess you to search web
> pages for a thing like this?
<fx: puts on "The Ideal Newbie" hat>
Just on the off chance. It takes about 30 seconds, and it's the least I
can do before bothering the 1,000 or so people on -questions.
I could've skipped this step. But I'd feel pretty stupid if I've bothered
lots of people with a question that I could have solved if I'd taken just
30 more seconds of my time.
<fx: takes hat off again>
> How would you have tackled searching for an answer to the problem of cron
> saying "Cannot fork"? I found nothing but my searching skills are clearly
> not up to the task yet. (Note: I found the answer by accident but still
> wonder how I should have found it)
Good question. I saw your original question, and didn't reply to it
because I couldn't answer it in anything more than general terms.
My train of thought goes thus:
'Cannot fork:'. Ok. fork() is a system call (this is something that
_I just know_, but I could have typed 'man -k fork').
A system call is failing for some reason. That's probably bad. Why
can this system call fail?
'man fork'. Ah, the errors section:
Fork() will fail and no child process will be created if:
[EAGAIN] The system-imposed limit on the total number of processes under
execution would be exceeded. The limit is given by the
sysctl(3) MIB variable KERN_MAXPROC. (The limit is actually one
less than this except for the super user).
[EAGAIN] The user is not the super user, and the system-imposed limit on
the total number of processes under execution by a single user
would be exceeded. The limit is given by the sysctl(3) MIB
variable KERN_MAXPROCPERUID.
[EAGAIN] The user is not the super user, and the soft resource limit
corresponding to the resource parameter RLIMIT_NPROC would be
exceeded (see getrlimit(2)).
[ENOMEM] There is insufficient swap space for the new process.
OK. That tells me why it might fail, but it doesn't tell me the specific
reason for failure. Do any of those look like they apply to me? How
many processes have I got running at the time? Does my machine have
enough memory? Am I running out of swap space? Can I try tuning one
of the sysctl(3) ('man 3 sysctl', then 'man 8 sysctl') parameters
and see if the problem goes away.
Hmm, try searching the web site for "cannot fork". . . no, that turns
up nothing. What about the mailing list? Nope, not a thing.
OK, a post to -questions is probably in order. Include the name of
the program that's failing, what time it's failing ("Ah ha, there's
a thought, am I running anything else that's load intensive at that
time, maybe out of the /etc/daily, weekly or monthly?"), how much
memory of got (since that seems to be a factor), the swap space and
the version of FreeBSD I'm running. Also include the user that's
running the program ('root' I think in this case) and the resource
limits for that user (the output of the 'limit' command).
That just about covers it, in a stream-of-conscious kind of way.
> Yeah, well that's just my opinion of the moment after a disk crash, an
> inadequate backup, and a long frustrating day. Try me again tomorrow :-)
Ah, one of those.
May I recommend an alcoholic drink? I'm quite partial to a Long Island
Iced Tea myself, 3 or 4 of those and I stop caring.
Right, I'm off to tackle some more of the handbook conversion project.
Have a good one.
N
--
Work: nik@iii.co.uk | FreeBSD + Perl + Apache
Rest: nik@nothing-going-on.demon.co.uk | Remind me again why we need
Play: nik@freebsd.org | Microsoft?
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-newbies" in the body of the message
help
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19980402183032.37825>