Bug 134086 - Problems running ipsec
Summary: Problems running ipsec
Keywords:
Status: CLOSED UPSTREAM
Alias: None
Product: Fedora
Classification: Fedora
Component: openswan
Version: rawhide
Hardware: i386
OS: Linux
medium
medium
Target Milestone: ---
Assignee: Harald Hoyer
QA Contact:
URL:
Whiteboard:
Depends On:
Blocks: FC3Target
TreeView+ depends on / blocked
 
Reported: 2004-09-29 15:27 UTC by Matthias Saou
Modified: 2007-11-30 22:10 UTC (History)
1 user (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed: 2004-10-14 11:11:00 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)

Description Matthias Saou 2004-09-29 15:27:23 UTC
Description of problem:
I'm trying to set up IPSec on Fedora Core 3 test2 and am having many
annoying issues :
- On one machine, I have the exact same
/etc/ipsec.d/examples/no_oe.conf file as on the others, but when I try
to run "service ipsec start", I get "/etc/init.d/ipsec:
(/etc/ipsec.d/examples/no_oe.conf, line 12) section header "
auto=ignore" has wrong number of fields (1) -- `start aborted", which
I really don't understand as all lines have "auto=ignore", that one
not having anything in particular.
- When running "service ipsec start", something gets either stuck or
not properly daemonized since a while loop with initlog calls is still
running after the service command has finished its execution :

/bin/sh -c while read line; do /sbin/initlog -s $line -f
${LOG_FACILITY:-daemon} -p ${LOG_PRIORITY:-error} -t ipsec__plutorun; done

- Last, the docs are really poor, messy and outdated (that's actually
more on the openswan side, I'll open another bug later on for that).

Version-Release number of selected component (if applicable):
ipsec-tools-0.3.3-1

How reproducible:
The running while loop with initlog is 100% reproducible on all 4
machines I've tested it on. The "wrong number of fields" problem is
100% reproducible on the only machine it's happening on, even after
"rpm -e ipsec-tools openswan && rm -rf /etc/ipsec.*" and reinstalling
both packages.

Steps to Reproduce:
1. Install ipsec-tools & openswan
2. service ipsec start
3. ps ax | grep initlog
  
Actual results:
/bin/sh -c while read line; do /sbin/initlog -s $line -f
${LOG_FACILITY:-daemon} -p ${LOG_PRIORITY:-error} -t ipsec__plutorun; done

Expected results:
Nothing.

Additional info:
I've already had a similar "wrong number of fields" error when I had a
config file with an error or bad indentation, but here I'm using the
default config files and didn't edit anything. The system was
installed with a minimal set of packages and ipsec-tools and openswan
were later added with yum, so maybe it's a missing runtime dependency
not pulled in by the packages? Just a thought.

Comment 1 Matthias Saou 2004-09-29 17:27:57 UTC
Sorry, this bug should ovbiouslt have been against openswan and not
ipsec-tools, changing that.

Comment 2 Paul Wouters 2004-10-03 19:31:46 UTC
Unfortunately, ipsec.conf's parser is due for an update (see
programs/starter). It is very sensitive to spaces and tabs (which are
not both 'white space').
You should have conn name, and on the next line have <tab>option, and
the end of the conn must have an empty line.
We've added a trailing space to the conn lines which seems to fix the
missing tab problem in CVS head.

We will look into the logging loop, but the fedora package changed a
lot of the logging functions, so I am not sure yet what code is at
fault here.

Comment 3 Matthias Saou 2004-10-04 11:56:23 UTC
Well, with the exact default examples/no_oe.conf, I was getting that
problem. Adding a white space at the end of every conn line seems to
have worked around it, weird. Thanks for the tip.

Comment 4 Ken Bantoft 2004-10-05 14:08:49 UTC
Fixed in Openswan 2.2.1 (released shortly) - changed the no_oe.conf.in
file in the distribution.


Note You need to log in before you can comment on or make changes to this bug.