We have always had issues with eLOG crashing intermittently...  I upgraded from Server 2003 to Server 2008 about a year ago to try to reduce the issues...  which really didn't help.

The service now seems to crash every week...  (getting worse)...  Is anyone seeing these issues in Windows servers?  Any suggestions??

Joe,

Run the command 
`limit' (if you are running tcsh) and `ulimit -a' if you are running bash
(these are command line programs that will tell you the resources you are
allowed to use.  Look at filesize, datasize, and stacksize - these need to be
8M or so)

You can increase these by typing, e.g.,

limit filesize unlimited
limit datasize unlimited
limit stacksize unlimited

(the equivalent command for bash are `ulimit -f unlimited', `ulimit -d
unlimited' and `ulimit -s unlimited', without the quotes)

which will increase the limits to the max imposed by the administrator.  
If that doesn't increase your limits, you will have to contact the admin to
have them raised system wide.

If you are the administrator/sole user, you can set them to the max in the file:

/etc/rc.common.

I have included the following at the end of the file:

##
# Enable coredumps if requested.
##
if [ "${COREDUMPS:=-NO-}" = "-YES-" ]; then
    ulimit -c unlimited
    ulimit -d unlimited
    ulimit -s unlimited
fi


and I have put this in /etc/hostconfig

COREDUMPS=-YES-

Hope this helps - if that doesn't fix it, sorry!

Sridhar

> I have been trying to build and run elog under Mac OS  X 10.2.6, without
sucess.  I 
> understand that this is not one of the 'supported' platforms, but I hope
that there is someone   
> here who might know the work-around to my problem.
> 
> I've tried building elog 2.2.5, 2.3.6, 2.3.7, and 2.3.8.  I've tried using
the compiler that comes 
> in the Mac OS X 10.2 box, as well as the gcc-based compiler supplied in
Apple's Dec 2002 
> developer update (gcc 3.1), and the June '03 update as well (gcc 3.3).
> 
> elog builds without error in all cases I tried.
> 
> When I run it, it will correctly serve the demo notebook.  However, when the
user clicks on the 
> test message to display it in full, there is a seg fault.
> 
> I suspect that the trouble might be with something defined in time.h.
> 
> I've run it in gdb, and the result is appended to this message.  This is for
gcc 3.3 and elog 
> 2.3.8.
> 
> 
> I would greatly appreciate any hints or suggestions.
> 
> Cheers, Joe
> 
> ---
> 
> [satsuma:~/Documents/elog_build/elog-2.3.8] jgiaime% gdb ./elogd
> GNU gdb 5.3-20030128 (Apple version gdb-282) (Fri Jun 13 03:33:07 GMT 2003)
> Copyright 2003 Free Software Foundation, Inc.
> GDB is free software, covered by the GNU General Public License, and you are
> welcome to change it and/or distribute copies of it under certain conditions.
> Type "show copying" to see the conditions.
> There is absolutely no warranty for GDB.  Type "show warranty" for details.
> This GDB was configured as "powerpc-apple-darwin".
> Reading symbols for shared libraries .. done
> (gdb) run
> Starting program: /Users/jgiaime/Documents/elog_build/elog-2.3.8/elogd 
> Reading symbols for shared libraries . done
> Indexing logbook "demo" ... ok
> Server listening on port 8080...
> 
> Program received signal EXC_BAD_ACCESS, Could not access memory.
> 0x00005870 in el_retrieve (lbs=0x1, message_id=50, date=0xbfff2500 "demo", 
> attr_list=0x0, attrib=0x365620, n_attr=1, text=0xbfff2280 "1", textsize=0x1, 
> in_reply_to=0x0, reply_to=0x0, attachment=0x0, encoding=0x0, locked_by=0x0)
at src/
> elogd.c:2775
> 2775    {
> (gdb) up
> #1  0x0001d9b4 in interprete (lbook=0x1 <Address 0x1 out of bounds>,
path=0xbfff2280 
> "1") at src/elogd.c:12279
> 12279       show_elog_message(lbs, dec_path, command);
> (gdb) up
> #2  0x0001db10 in decode_get (logbook=0xbffff340 "demo", string=0x3a300a
"1") at src/
> elogd.c:12325
> 12325     interprete(logbook, path);
> (gdb) up
> #3  0x00020080 in server_loop (tcp_port=836736, daemon=-1073732640) at src/
> elogd.c:13398
> 13398             decode_get(logbook, p);
> (gdb) up
> #4  0x00021134 in main (argc=1, argv=0xbffff850) at src/elogd.c:13832
> 13832     server_loop(tcp_port, daemon);
> (gdb) up
> Initial frame selected; you cannot go up.
> (gdb) 

 1. It appears that some times find_thread_head is called with message references that do not exist. That is not good.

I put in a little check like this  before seeing if the message has an "in_reply_to" reference:

The line:

if (lbs->el_index[i].in_reply_to)

becomes:

if (i < *lbs->n_el_index && lbs->el_index[i].in_reply_to)
 

2. The trouble started when I deleted a message in the middle of a thread, which left the thread badly "connected" (references to a deleted message).

3. Also, when a thread is badly connected, it is a problem moving messages to a different logbook. ELOG complains that it cannot access the message (with the invalid reference). But ELOG should ignore it, since the message was deleted.

Soren

 It would be nice to have this corrected. The problem occurs when you select (read) a message which refers to another message via "In-reply-to", and this message does not exist.

Soren

Hi,

First of all, thanks for all the great work, we all thank you here at SESAME synchrotron 

I am trying to set a conditional attribute where selecting one option in that attribute should trigger multiple outcomes as shown below:

Options Type = Fault{1_1&1_2}, Test{2_1&2_2}, Report{2_1&2_2}

{1_1} Options Fault Status = Open{a}, Fixed{b}
{2_1} Preset Fault Status = NA
{1_2} Preset Icon = attn.png
{2_2} Preset Icon = icon1.gif

I tries adding "&" as shown but that didn't work, my goal is to have 1_1 and 1_2 triggered whenever I select "Fault" as type, how can I do that?

Thanks in advance.

Cheers,

Sofian.

Thanks Andreas,

Works great.

I tried reading the documentation on this but I was not able to deduce this solution.

Thanks a lot again.

Thank you Stefan as well