Demo Discussion
Forum Config Examples Contributions Vulnerabilities
  Contributions to ELOG, Page 4 of 6  Not logged in ELOG logo
New entries since:Thu Jan 1 01:00:00 1970
Entry   Code change for LDAP authentication, posted by Maurat on Mon Mar 11 09:28:15 2019 auth.c

Hi,

I had to change code to authenticate users in my organization's LDAP directory. Indeed, accounts are distributed under several organizational units in my LDAP directory.

The current version of the code can't authenticate accounts when these are in different organizational units. Hence my contribution.

I Use a read account to request LDAP to locate the account that has logged in (with e-mail address in the search filter).

I get the number of LDAP entries. If I have one entry then I call ldap_get_dn function to get the DN account and then I call ldap_simple_bind_s using the account's DN and password to perform LDAP authentication.

I changed configuration file elogd.cfg. I added two parameters:

LDAP DN user = <DN read account>

LDAP PW user = <password read account>

I changed code auth.c too (see attached file)

I had to change Makefile. I added a call to lber library

ifdef USE_LDAP
ifneq ($(USE_LDAP),0)
CFLAGS += -DHAVE_LDAP
LIBS += -lldap -llber
endif
endif

Have good day

 

 

 

 

 

 

    Reply   Re: Custom input forms implementation, posted by John on Sat Jun 15 06:13:07 2019 

I have been trying to get my head around this application module. I assume that after the input is done on this example (ShiftCheck), if goes into the  usual file system directorys for storage? Or is it (or can it) goto another db like sql, flat, etc.?  I also assume that the 'normal' Elog screens we see for input (and output), would be a completely seperate module that you have for (ShiftCheck).. but we do not see them here (as one of the attachments)? I am asking these questions because I am trying to recreate this (type) of input/output system for users, so I would like to know how the 'whole picture' is done with your example here. Thanx again.

Stefan Ritt wrote:

Dear ELOG users,

starting with SVN revision 2328, custom input forms are implemented. This allows application specific formats for check lists etc. In our specific case we had to implement a shift check list, which was quite long. Furthermore the check list should be optimized for an iPad, which we take in the field and record various checks and readings (in our case some gas pressure gauges at the PSI particle accelerator). Since the standard ELOG interface was too inflexible, a completely hand-written form was needed. The form can be activated by the new configuration options Custom New Form, Custom Edit Form and Custom Display Form, one for a new entry, an entry to edit and and entry to display. In our case we used the same form for all three cases. This is how the shift check list looks under the Safari Browser on a PC:

Capture.png

And here is how it looks on the iPad:

IMAG0036.jpg

Each section can be collapsed and expanded (blue arrows at the left), and various internal checks are made before the check list can be submitted.

Implementing such forms is however more something for the advanced user, since you have to hand-write HTML with CSS and JavaScript code. It can then however be a powerful method for check lists. Please find in the attachments the elogd.cfg configuration for that logbook and the shiftcheck.html source code file. It is a bit complicated since the page is a static page, elogd just serves it from the file. This requires all the dynamic functions to be implemented inside the HTML file with JavaScript. To display an entry for example, the JavaScript loads the raw data with the "?cmd=Download" command and the populates the form fields. The collapsing and expanding is done by using CSS properties. The integrated style sheet was optimized for the rendering on an iPad. Rather large fonts were chosen so that the items can be checked easily with your finger tips. Various parameters are sent between the browser and the elogd program via hidden fields and cookies. So only something for experts! But if you go through the effort and hand-write the form, it can be very handy. Note that you have to upgrade to SVN revision 2328 for the three new options.

 

 

Entry   Steps for securing Elog using SSL and Apache, posted by damon nettles on Wed Nov 24 23:45:19 2004 elogredirect.conf
Everything in this guide was done on a full install of Fedora Core 3 running
Apache 2.0. If you are using an older version of Apache some of this may not
work, so I recommend upgrading. Also, on different Linux distributions, some
of the paths may be different.


The goal here is to get Elog set up under Secure Socket Layers, so that
communication both ways is encrypted.  This will cover any password
transactions so nothing gets sent over the web in the clear.

The previous method of securing the Elog, which involved using stunnel, is
out of date. A better way to go is to use the Elog in conjunction with
Apache. The Apache method leverages all the research and development that's
gone into providing secure sockets for Apache, and removes the need for any
serious reinventing of the wheel.


We begin with a web server running on port 80 and an Elog server running on
port 8080.


Making Certificates:
It's necessary to generate some secure certificates to be issued to anyone
who attempts to access the securesite.
A guide to making the certificates can be found at:

http://slacksite.com/apache/certificate.html

So, following the steps in the article:
   openssl genrsa -des3 -rand file1:file2:file3:file4:file5 -out\
   server.key 1024 
where the \ is merely an indicator that the command wouldn't fit on a line
here.  The fileN references are sources of random information to help the
random number seed be more random.  I merely used some personal text files
that were zipped up, as suggested in the page.

   openssl rsa -in server.key -out server.pem

Removes the RSA encryption from the key, to make it easier for the Apache
server to deal with it.

   openssl req -new -key server.key -out server.csr

Starts a line of questioning about us as a certificate issuing entity.
Answer with reasonable values.

  openssl x509 -req -days 60 -in server.csr -signkey server.key -  
  out\ 
  server.crt

After this move the server.pem, server.crt, and server.csr to the
appropriate directories under /etc/httpd/conf/ .  The extensions explain
which directory to put them in, with the exception that server.pem ended up
in etc/httpd/conf/ssl.key/ .


In the elogd.cfg file, change the port to 8079, and set the URL to
"https://your.host.name/" .  Restarting the Elog daemon now leaves us with
Elog listening to port 8079 instead of port 8080.


The rest of the story is in the "elogredirect.conf" file attached to this
post, but here are the highlights.

Create a virtual host dealing with SSL that listens to port 443 (the ssl
port), and acts as a proxy for port 8079 (where Elog is listening).  This
allows Apache to act as an SSL handler for Elog by handing off any access at
https://your.host.name/ to the Elog server.  The firewall then can keep out
any direct attempts to access port 8079, so that the only thing that can
reach the Elog server is stuff talking to 8079 on the local side of the
firewall (which pretty much means just the Apache proxy).  I recommend
Firestarter for the firewall config by the way, it's a real lifesaver.

http://firestarter.sourceforge.net/

This covers the SSL portion of the story, and by doing the redirection
inside the port 443 virtual host, instead of from the port 80 webpage as
before, you can avoid any path overlap.

As was the case for us, you may have links in older Elog posts, e-mails, or
web pages that point to specific Elog posts. If you have been using Elog for
some time and never bothered with the SSL stuff, the links most likely look
something like
http://your.host.name:8080/yourlogbook/postnumber. 

To cover legacy support for calls on port 8080, you can  create another
virtual host listening to port 8080.  This host's job is to take any
incoming URL calls on "http://your.host.name:8080/a_directory" and
translate them into calls on "https://your.host.name/the_same_directory" .
This means that any attempt to contact the Elog on port 8080 will get
answered by an Apache virtual host that redirects the client through the
Apache SSL virtual host described above. See the conf file for the details.

So in the end, the firewall is set to only allow through ports 80, 443, and
8080.  Port 80 handles the normal webpage access stuff.  Port 443
exclusively handles the SSL port for the Elog daemon, and port 8080
exclusively handles the redirect for the legacy Elog calls.

Implementation of this setup on another system should be pretty
straightforward.  Apache's config file is at /etc/httpd/conf/httpd.conf ,
and it also loads any *.conf files in /etc/httpd/conf.d/ .  So its a pretty
simple case of just dropping elogredirect.conf into /etc/httpd/conf.d/ and
restarting the Apache server.  Of course the necessary changes to elogd.cfg
have to be made and that server restarted as well.  The firewall, too, needs
to be setup to secure the whole deal. Note that the elogredirect.conf file
needs to be edited for your specific setup (changing the instances of 
"your.host.name" to whatever your server is, and also putting in the
administrator e-mail address where it is noted).


This work was done by Jonathan Hanson and Damon Nettles in the Gravity Lab
at Louisiana State University. You can see our Elog at
https://sam.phys.lsu.edu/elog .

If you have any questions or comments send them to
nettles@phgrav.phys.lsu.edu .
Entry   [New Feature]: JS calendar filter , posted by Emiliano Gabrielli on Thu Jul 28 18:42:48 2005 elog_calfilter.pngelog_calendar_filter-1.0.1.tar.gz
NOTE:
it seems that for some strange reason it slows down elog very much when and only when elog is stunneled over ssl!


The attached tarball contains a full featured JavaScript allowing everyone to use a JS calendar (no popup windows, just JS and CSS!) to perform date-based filtering actions. An uncompressed version and the tarball for the last cvs rvision of this script is available at www.SuperAlberT.it

You can browse the ChangeLog here.

A simple way to use it is to uncompress the wall package into the elog data dir, under the "scripts" subdir.

the you can add the following code to your elog.cfg:
Filter Menu text = scripts/calendar_filter/calendar_filter.html

note that this configuration parameter is available in elog starting from
Revision 1.732 2005/07/29

and automagically have the new calendar_filter icon showed Smile

Entry   Solaris 10 SMF/Greenline management manifest for ELog, posted by Peter Eriksson on Wed Sep 7 16:52:30 2005 elog.xml
Please find enclosed as an attachment a Solaris 10 SMF/Greenline manifest that can be used to manage ELog.
(If you don't know what it is - it replaces init.d/cron/inittab and more stuff)
Entry   Compiling elogd.c on HP-UX 64 bit, posted by Peter Rienstra on Wed Jul 11 11:13:16 2007 elogdhpux64.c.gz
We succeeded in compiling and running elogd (elog-2.6.5) on HP-UX 64 bit Itanium platform (HP-UX B.11.23 U ia64).

The main problem was we got a core dump after starting elogd. The cause was that the memory has be allocated with a 4 byte boundary. This could be the case on other 64 bit platforms as well. A colleague of mine (Sander Notting) found the solution.

Unzip and untar the zip file (elog-latest.tar.gz)
Go to the src directory (elog-2.6.5/src)

Edit elogd.c

Replace all:

show_selection_page(NULL); => show_selection_page();
seteuid => setuid
setegid => setgid

On line 564:
void *buffer => char *buffer


Line 645, add the text in bold:

void *xmalloc(size_t bytes)
{
char *temp;

/* Align buffer on 4 byte boundery for HP UX and other 64 bit systems to prevent Bus error(core dump)*/
if (bytes & 3)
bytes += 4 - (bytes & 3);


temp = (char *) malloc(bytes + 12);


After that compile:

cc -w -c -o regex.o regex.c
cc -w -c -o mxml.o ../../mxml/mxml.c
cc -w -c -o strlcpy.o ../../mxml/strlcpy.c
cc -I../../mxml -o elogd elogd.c regex.o mxml.o strlcpy.o

We didn't try to run elogd under root yet.
    Reply   Re: Compiling elogd.c on HP-UX 64 bit, posted by Stefan Ritt on Fri Jul 13 12:36:45 2007 
I applied most of your patches to the elog source code, SVN revision 1885. The only missing piece has to do with seteuid/setuid. I definitively need seteuid for linux, because elogd might be started under root, then it falls back to an optional elog user. But when it stops, it has to restore the original root user in order to delete the PID file (/var/run/elogd.pid) which was created under root. If seteuid does not exist under HP-UX, you should add something like
#ifdef HP-UX
  setuid(...)
#else
  seteuid(...)
#endif

Probably the HP-UX has to be something else, but I cannot test this since I don't have such an OS here. Once you get this working I can put it into the standard distribution.
    Reply   Re: Compiling elogd.c on HP-UX 64 bit, posted by Peter Rienstra on Mon Jul 16 15:27:08 2007 man_setuid.txtman_setreuid.txtman_setresuid.txt
Stefan,

First I want to say I really like your program. We work in a small group of 5 database administrators, and this is exactly what we need to inform each other. Elog is simple but very functional, so thanks!


My problem is that I don't have root access to the HP-UX machines. We don't run elogd as root, so I wasn't really interested in the seteuid functionality, I just wanted to compile and run the program.

HP-UX doesn't have the "seteuid" and "setegid" functions. But there are "setuid+setgid", "setreuid+setregid" and "setresuid+setresgid" functions available. I'm not sure which one is the best to use. I uploaded the manpages as attachment. I hope this will help you.

If you want I can do a compile and run test on HP-UX with your altered source code. But I can't do a test with "root".




Stefan Ritt wrote:
I applied most of your patches to the elog source code, SVN revision 1885. The only missing piece has to do with seteuid/setuid. I definitively need seteuid for linux, because elogd might be started under root, then it falls back to an optional elog user. But when it stops, it has to restore the original root user in order to delete the PID file (/var/run/elogd.pid) which was created under root. If seteuid does not exist under HP-UX, you should add something like
#ifdef HP-UX
  setuid(...)
#else
  seteuid(...)
#endif

Probably the HP-UX has to be something else, but I cannot test this since I don't have such an OS here. Once you get this working I can put it into the standard distribution.
    Reply   Re: Compiling elogd.c on HP-UX 64 bit, posted by Stefan Ritt on Mon Jul 16 15:57:47 2007 

Peter Rienstra wrote:
HP-UX doesn't have the "seteuid" and "setegid" functions. But there are "setuid+setgid", "setreuid+setregid" and "setresuid+setresgid" functions available.


I had a look and found that setreuid/setregid is also available under Linux, so I use those functions instead, which should also work on HP-UX. Can you check revision 1888 (http://savannah.psi.ch/viewcvs/trunk/src/elogd.c?root=elog&rev=1888), compile it and see if you can run it at least under your non-root account.
    Reply   Re: Compiling elogd.c on HP-UX 64 bit, posted by Peter Rienstra on Mon Jul 16 16:43:07 2007 

Stefan Ritt wrote:
Can you check revision 1888 (http://savannah.psi.ch/viewcvs/trunk/src/elogd.c?root=elog&rev=1888), compile it and see if you can run it at least under your non-root account.


I downloaded revision 1888. There were no problems compiling it. It's running on the HP-UX system now and everything seems to work fine. Smile
ELOG V3.1.5-3fb85fa6