Dear all,

I have elog version 3.12-bd75964 installed on an Ubuntu OS. We started working with it yesterday and today I was informed that the users could not connect to the elog. When I tried it was also not possible to get to the elog website. After restarting the Ubuntu machine everything was back to normal.

Is there a log I can check to find out why the elog stopped working?

Regards,

John

I don't know if you can get elog to generate a log file - check the documentation, I don't bother.  But I do have some experience with this matter.

There are two related circumstances I know of that can arise which will cause elog to crash.

As you are probably aware, the entries are threaded.  I find the main problem is if you move a long thread - I forget the number but over 30 entries - from one log book to another, it will copy across fine, but the deletion of the thread from the source logbook will crash elog *before* the entire thread has been deleted: cause 1.

If you then restart elog, and happen to access the partially deleted thread, elog will run into a loop.  You won't be able to access elog, and you will see it is burning up CPU time, and it probably will eventually crash, but normally I find what the process number is and kill it (with "kill -9 [process no]").  This is cause 2.  The reason is that elog is looking for an entry number in the thread that no longer exists (because elog has already deleted it).

There is a moderately convoluted way I have devised to completely delete the partially deleted thread, needing access to the yymmdda.log files.  Cause 2 can also occur if someone manually edits yymmdda.log files with insufficient care (that includes me on occasion).  

I have a crontab set-up for root:

crontab -e

* * * * *       if ! ps -C elogd >/dev/null;then /sbin/service elogd restart;fi

The script just checks if elogd is still running and if not, it'll restart it. We run Centos, but I'm sure it'll be easy to adapt for Ubuntu (I don't know much about Debian).

This will only help if elogd really crashed; in case it is still running at 100% cpu load this won't help.

In our case (~30 logbooks, > 100 entries per day, ~ hundred users) elogd is crashing about once a week. With the help of this script it means 1 minute downtime a week: that's acceptable.

See https://midas.psi.ch/elog/config.html on how to use a logfile with elogd. Here's the relevant excerpt:

• Logfile = <file>
  This option specifies a filename which logs all login/logout activities and successful user connections for logbooks with user level access. The the logging level (see below) is larger than 1, also read and write accesses can be logged.
• Logging level = 1 | 2 | 3
  Specifies the logging level. The higher this value, the more information is logged. Default is 2:
  • 1: Log only logins and logouts
  • 2: Log also write accesses
  • 3: Log also read accesses

We have elog-2.7.5-1.i386 on SL 5

If I create an elog entry using the web interface, and include an inline image,  email is sent with a zero-length named attachment - the MIME header is present, but no content.

In the config file, Email Format = 47, though I also tried with format = 63.

Is this a bug that was fixed in a later version, or a configuration error (or a new bug) ?

Probably us not having ImageMagick installed. elog was able to attach pdf's, xpm's and xbm's to email, but not jpeg's or png's, though they inlined OK in HTML on the server.

It seems OK, I think, after installing ImageMagick and restarting.

Hi,

As mentionned before, we happen to fail to receive email messages related to updates on elog entries at our site. My understanding is that the SMTP header Message-Id MUST be unique for each email message. Whereas all elogd email messages get something like <logbook>-<entryId>@<domain>. See source code. For this header to become unique, there should be a random part in it.

Having the same Message-Id in multiple email messages results in only the first one being delivered on some email systems.

Moreover, elogd sets the In-Reply-To: header in the same manner (<logbook>-<entryId>@<domain>). Which is incorrect because this header relates to email messages, not elog entries, and should contain the email Message-Id of the email message to which it replies, itself handled by the email messaing system. But elogd hasn't received any email messsage in the first place. So I believe this header should simply be dropped.

I think I can provide a pull request on bitbucket for the Message-Id issue, and probably also for the In-Reply-To: if you decide it can be removed.

Cheers

A pull request would be highy appreciated, because you can then test it thoroughly on your side. Adding a random number to the message id is simple. "Reply-to" indeed does not make sense since elog cannot receive emails. Most sites use a generic "noreply@<domain>" to indicate to the user that a reply does not make sense. I guess the "Reply-to" does not have to be unique, right?

Pull-request posted. Cheers.

Pull-request merged.

Hello all,

I would like to configure the email that is sent to notify users, when an entry is submitted:

- Is it possible to attach to the mail the files attached to the entry ?

- Is it possible to configure a string (with entry keywords) that will appear in the mail content instead of the default content ?

Thank you!

Christine

Please look in the manual under "Email Format" and "Use Email Subject".

Hello,

I'm sorry but that's a part of the manual I've read many times, and I'm not interested in changing the subject of the email, but only body and attachements.

I guess that if I didn't find these answers in the manual, it's because the answer for both questions is no ?

Thank you

I'm sorry but you are only half right. The mail contenct cannot be changed. But if your email notifications are sent as HTML, they contain a 1:1 copy of the elog entry, which does not make sense to change. For the attachment to be attached to the email, you need the flags of "Email Format". The default there is 63, which means that elog attachments ARE sent as mail attachments.

hi

I have an elogsystem for documenting emergency services. I have a starttime and an endtime. Is it possible to calculate the diff between the 2 timestamps and convert it to minutes?

Attributes = Bereitschafts Kalender Woche, Bereitschafts Woche, Einsatzbeginn, Einsatzende, Einsatzzeit, Telefon Nummer, Station, Status, Problem 
Options Status= Offen{1}, Erledigt{2}, Geschlossen{3}

......

Type Einsatzbeginn = datetime
Type Einsatzende = datetime

....

Subst Einsatzzeit  = $shell(Einsatzende - Einsatzbeginn) 

regards

/alex

I found the solution in the forum :-)

subst Einsatzzeit = $shell(/usr/local/elog/diff.sh \"$Einsatzbeginn\" \"$Einsatzende\")

and diff.sh is

#!/bin/sh
#set -x
s1=$(date -d "$1" +%s)
s2=$(date -d "$2" +%s)
d=$(echo "($s2-$s1)/60"|bc -lq)
printf "%.0f Minuten" $d
 

is ist possible that subst only works for the initial create of a logbook entry? when I change/edit one of the parameter "Einsatzbegin" or "Einsatzende" the "Einsatzzeit" is not changed.

regards

/alex

There is the "subst on edit = ..." option you should use in addition.

Stefan

thx!

It is possible to change the justification for a column/attribute in the list mode (left-aligned, centered, right-aligned)?

Right now Elog treats each colum of a list as centered,

thank you

You can do that chaning the CSS of the underlying theme. But you need some experience with CSS.

I need to set text-align:left but only for an attribute in the List mode, For example, say that FirstName column must be left justified in list view. What is the name of the cell content in the default elog.css?

You might tell your browser to show you the source code of an ELOG page in list mode. There you'll find the class name you're looking for.

Cheers, Andreas

The CSS class for the list display table "listframe". If you want the second column to be left-aligned, you put following into elog.css:

.listframe td:nth-child(2) {
   text-align:left;
}

that should do the trick. Replace the "2" with the column you would like to change.

Stefan

Thank you Stefan: it worked!

Now I am wondering if there is some syntax to force this behaviour at the attribute level in elog.cfg, without the need to modify the css, I mean simply based on the attribute name (not its content). The syntax Style <attribute> <value> = <style> can modify the style but it applies to the whole line in the list, so all the cells in that line are affected. Also it apply *only if* the content of the cell equals <value>.

In other words, an overridden version for Style: Style <attribute> = <style> that will be applied *regardless* of the cell content _AND_ only to the cell specified in <attribute>

What do you think?

This is currently not possible.

/Stefan

Hello,

I wonder if it's normal that in the sources files, new created mxml directory is empty.

In commit # a6e5962 4 files are deleted but not moved:

- src/mxml.c
- src/mxml.h
- src/strlcpy.c
- src/strlcpy.h 

Thank you

The mxml package has been converted into a submodule in the git repository. At http://midas.psi.ch/elog/download.html you see that you need a "--recursive" flag to clone it. If you have already checked out the sources, you need a

git submodule update --init --recursive

/Stefan

I've created an elog where users can then enter links to a file: for example they would enter:

<a href="file:///M:/Data/Inspection%20Reports/ABC123.doc">ABC123</a>

If they are running this on their desktops where everyones M drive is mapped to the same location, not a problem, the file opens perfectly.

The problem we have is some people work off of site and hence they run everything from a memory stick (USB drive), then once back in the office, copy the memory stick back on to the network.

Unfortunately when off of site the href link M: is no longer valid as the M drive doesn't exist.

Hence i've been trying to open a local file with the following syntax: 

<a href="file:///../ABC123.doc">ABC123</a>

It just won't open

Hence in my frustration i basically copied the word file ABC123.doc into every single directory on the memory stick thinking surely this must work, it must find it somewhere, unfortunately not.

What am i doing wrong?

Linking to external files via "file:///..." is a bad idea, since only a pointer is stored in the elog. So if someone puts somehing on M: and the drive is removed, the file is physically gone. No way to retrieve it. Instead of puttting links to files, users should get trained to ATTACH the files to elog cntries (V3 of elog has a "Drop attachments here..." section. Only then the files are physically transferred into elog and accessible independed of your external memory stick.

Stefan

Hi Stefan

Unfortunately this would become impractical for us as many of the test data files which we'd need to link too are Gb's in size, some over 100Gb, hence only the files needed to do a say an inspection of the equipment would be copied onto memory sticks. Currently we use an excel speadsheet to store hyperlinks to all of the files / directories we need - unfortunately some people forget to update this and it was asked could we do it using elog as people are familiar with this wonderful peice of software.

Hence why i'm now investigating - so is it actually posible?

Kind regards

Terry 

If a file is on a memorystick and the memorystick is gone, how could elog bring it back? What you ask for is magic, sorry for that. Also if you move files outiside elog from one folder (drive) to anohter. How shoudl elog keep track of these files if they are outside? If your colleague moves some piece of equipment from A to B, without telling you, how can you guess where B is?

Hi Stefan,

Obviously it's not magic but my inability to explain. I've attached a quick jpg of the file structure we use. So for every project we just repeat the file structure. It then becomes a browser based interface for our engineers, inspectors, project managers etc. so via the network they can access the data anywhere onsite.

Elog is used in an iframe on the manrec page. Here we record several logs (a record of everything we do, a record of test equipment, tools used, harnesses, connectors made and disconnected, etc. and what i'm now trying to achieve a document record / check list).

Now in the data directory we store hundreds of documents from procedures, circuit diagrams, mechanical drawings, parts lists, test reports and data, inspection reports, down to consumables like what type of cloth and liquid was used to wipe down a cable.

So the bulk of documents can fit on to an 8G memory stick, the only section that doesn't fit is the test data.  So when we need to move off site, or supply this data to a subcontractor we copy the contents (minus the test data) on to a memory stick, when the memory stick is returned we simply synchronise it back to the network. So the only links that would fail to work off site are links to test data - which we are not worried about.

Now, any new person working on the project needs to review the check list (hopefully the new elog if i can get it to work) part of this check list  / document record could say something like - carry out inspection, refer to drawing <href link> now this is where the problem occurs, as when the files are on the site based machine, we know the file structure is on the M:Drive; however when the memory stick is now plugged into a subcontractors machine / laptop we have no control of the drive allocation, so how can i get the link to the drawing to work?

When they finish the job required, they upload the data to the memory stick and state where.  ie inspection carried out - results passed see report "xyz" <href link>

I hope this now makes more sense

Kind regards

Terry

Sure, we (Andreas and I) get your point. But you say "... we have no control of the device allocation". If you have no control, how could elog have control? As Adreas said, the file:// access is even done inside your browser, thus bypassing elog, so you have to tell your browser where to find the files. I cannot say more than what Andreas said in elog:68562

Stefan

Hi Terry,