Hi,

I installed ELOG on other machine for making automatic backups. I copied the .cfg file to the slave and edited the .cfg file on the original machine (mirror server entries).
(automatic configuration with -C didn't work somehow - maybe a firewall problem)

Then I tried to synchronize manually, which works great.
I set the mirror cron = 0 7-18 * * 1-5, as written as example in the documents, but nothing happens at full hours.

There's something else strange: I can't access "https://lx0.mydomain.org/" or "http://lx1.mydomain.org:8082/" directly. The browser just keeps loading - showing an empty page (and an error message after a long while). I have to add the name of the logbook to the link. Why is that?
If I set different paths in the [global HBAR] and [global SMI] areas (like URL = https://lx0.mydomain.org/hbar/), I can access the links above (giving me the overview page as expected), but I can't access the logbooks itself anymore.

However, that's not a big problem (I can access the logbooks via direct link and see all the others in the same group and switch to them), but the backup thing is.

So, main machine is lx0.mydomain.org and the backup machine is lx1...

Attached are my config files.
I deleted many of the logbook specific entries, since they aren't important and just look like the first one.

Hope anyone can help,
Roland.

The synchronize feature is totally worthless to me.

First of all the automatic backup doesn't work - and nobody seems to know why - and pressing the synchronize button by hand from time to time also doesn't work if the entry is too big:

Error sending local entry: Error transmitting message

So, copying the files manually helps, but for this I don't need a "fancy" synchronize feature.

Hello everybody,

we had to setup a new server for our elog some time ago because the old one crashed. We are now running version ELOG V3.1.1-0767eb0.

In the previous installation we had these nice buttons for file upload, time stamp and equation editor, which are also available here in the online version.

I already found that in the installation directory ("/usr/local/elog/") there exists a directory "scripts" containing a file "ckeditor-config.js", which also contains the lines:
"config.extraPlugins = 'timestamp,dndfiles,eqneditor,fileupload';" and "{ name: 'insert', items: [ 'Image', 'Table', 'HorizontalRule', 'Smiley', 'SpecialChar', 'PageBreak', 'FileUpload', 'Timestamp', 'EqnEditor' ] },"

These plugins seem to be available in the subdirectory "scripts/ckeditor/plugins/". But why are they not loaded? Is it just a version issue? Or something else?

It would be nice to get a hint.

Best regards,

Roland

Dear Stefan,

thanks for your hints. Meanwhile we already solved the problem. I think it was caused by the recovery from a backup version after the server crash ... and the retirement of the person who did the previous installation.

We found the following things to be faulty:

1. Somebody linked the "elog/scripts/ckeditor"-directory to "../../javascript/ckeditor" (maybe an independent installation of ckeditor). Due to this the relative path in "ckeditor/config.js" to "../ckeditor-config.js" could not be correctly resolved to be "elog/scripts/ckeditor-config.js". Now we put the absolute path.
2. Furthermore the owner of all files in the elog directory was set to root:root with read access only for root. We changed them now to root:elog and gave read acess for everybody.

Now everything works as expected.

I'm sorry for inconvenience,

Thanks and best regards,

Roland

> Basically im running elog on a redhat 8 box with stunnel already installed 
> as part of OpenSSL, on the server i ran the command specified in the 
> instructions
> 
> stunnel -d 443 -r 172.16.24.108:81

I haven't tried `stunnel` with elog, but as far as I read from various
sources, stunnel requires you to create an SSL certificate. In attachment,
you'll find a sample case for stunnel and smtp (port 25) + pop3 (port 110).
Hope this helps.

Reference: http://www.tldp.org/linuxfocus/common/sart/index.html

Ooops, I destroyed the original post, now how can we fix this :)

> Here is the error:
> 
> Cannot find any ??????.log file in this directory.

`elconv' is just a simple tool to convert pre 2.0 log files to the current
format. (Well, after all these confusions, dropping it from the future
releases won't be surprising -at least for me- ;) You might try some
well-known text-processing tools ie. preferably awk or perl to make the
required conversions. But I think using comma or tab as the delimiter is not
suitable due to the content of logs. If all you want is importing them to
excel/word, you should play with the delimiter options of mentioned programs
before trying a solution.

Seems a stack problem.  I don't have an OS X box to validate my guess.  I've
made a Google search with the keywords: EXC_BAD_ACCESS stack "os x"

Perhaps the following reply [1] might help...

  How big are they? You're probably running into the default stack size 
  limit, which is 512 KB. Try `unlimit stacksize` in your shell before 
  running, which will give you 65536 KB per stack.

[1] http://zerowing.idsoftware.com/archives/gtkrad-macos/2003-January/000453.html