Handbook Jail Chapter rewrite available for critique

Alejandro Imass aimass at yabarana.com
Thu Mar 21 15:21:32 UTC 2013

On Thu, Mar 21, 2013 at 3:35 AM, Ian Smith <smithi at nimnet.asn.au> wrote:
> On Tue, 19 Mar 2013 17:53:30 +0100, Dirk Engling wrote:
>  > On 18.03.13 20:16, sib at tormail.org wrote:
>  >
>  > > to configure things themselves. In my experience, ezjail is a much better
>  > > solution. I also see that you are the maintainer/author of qjail and like
>  > > to shovel your opinion as the only solution, both in this "rewrite" and
>  > > all over the FreeBSD forums.


>  >
>  > # Copyright  2010,  Qjail project. All rights reserved.
>  >
>  > offensive. I am usually quite open with the license of my software,
>  > beerware is as permissive as it gets. I just can not take some script
>  > kiddie right out copying my code verbatim and selling it as his, not
>  > even acknowledging me as the original author.
>  >
>  > Anyone here with suggestions how to properly react to this kind of "fork"?
> Yes.  Publicity.  Making sure the FreeBSD community gets to finds out.


> To that end I'm cross-posting this to -questions, where Mr Barbish has
> also posted about his proposed "rewrite" of Chapter 16 of the Handbook,
> which is nothing but a huge and poorly written manual for 'the qjail
> way', with its peculiar assumptions and unique "jailcell" terminology.
> "Fourth Generation", no less!


Thank you Ian for cross-posting here.

The first thing I did when I got the new chapter for review was search
for the work EzJail and I was curious as to why EzJail is not
mentioned anywhere in this new proposal and why it isn't mentioned in
the current handbook either under in section "16.5.2 High-Level
Administrative Tools in the FreeBSD Ports Collection". If there is
__any__ tool that should be mentioned in the jails chapter it is
EzJail because it's really easy to use and does a damn good job.

We've been using it in production __extensively__ since about 2010 and
the one and only issue we've had was probably related to some sort of
border-line bug with nullfs which has never happened since. We
currently run half a dozen servers with anywhere from 12 to 24 jails
each and we've only had a single isolated incident and it wasn't even
related directly to EzJail. We use flavours extensively and constantly
derive jails from others and move jails between servers, much like if
we were using VMWare; it's that easy, or easier, and works every time.

NOW some things start to make sense to me, when I posted a problem
with EzJail here last year that very few people, if any, knew what I
was talking about. An how could they? if it's not mentioned anywhere
in the handbook or that jail man page(s).

In fact, looking back at this thread[1] I can see that great deal of
misunderstanding an unnecessary confusion could have been that the
term "EzJail" meant nothing to most people commenting on the thread.
When I commented the problem to Dirk he immediately recognized that it
could have been a problem with nullfs and so did "jb"[2], who not only
immediately thought of nulls, but actually found some bugs that were
very similar to my situation[3], and which is BTW still open AFAICT.

Anyway, the point I'm trying to make is that it seems quite odd that
EzJail is not very publicized and I would like to see it prominently
mentioned in the handbook and man pages as a great tool for Jail


Alejandro Imass

[1] http://lists.freebsd.org/pipermail/freebsd-questions/2012-April/240468.html
[2] http://lists.freebsd.org/pipermail/freebsd-questions/2012-April/240566.html
[3] PR#147420

More information about the freebsd-jail mailing list