Has anyone had issues with having too many files open? I'll setup my server and let it go but after a while, I end up with a lot of "cannot create socket: Too many open files" errors being reported.  I have a sync to another e-log going which I suspect is part of the cause since that e-log server hasn't had this issue. I suspect that there are files being opened, going into some return loop code and then never getting closed. I'm not a C programmer but I see lines like :

fh = open(tmp_filename, O_RDONLY);
      if (fh > 0) {
         read(fh, result, size - 1);
         close(fh);
      }

      /* remove temporary file */
      remove(tmp_filename);

This looks like it opens the file but unless the remove function closes the file, it will remain open even through the file has been deleted. Maybe this isn't the correct behaviour of 'remove' and I am mistaken?

There are also parts like :

 fh = open(textfile, O_RDONLY | O_BINARY);
      if (fh < 0) {
         printf("Message file \"%s\" does not exist.\n", textfile);
         return 1;
      }

      size = (INT) lseek(fh, 0, SEEK_END);
      lseek(fh, 0, SEEK_SET);

      if (size > (INT) (sizeof(text) - 1)) {
         printf("Message file \"%s\" is too long (%zd bytes max).\n", textfile, sizeof(text));
         return 1;
      }
This looks like for the second error, it will complain that the file is too long, return an error message but not close the file and would leave it open. Is this a reasonable avenue to pursue or am I mis-reading the code?   Thanks.

There is a python API to access ELOG via HTTP: https://github.com/paulscherrerinstitute/py_elog

EDIT: forget the tip below. Instead just call script files: inline scripting in the ELOG config shows very strange behavior. Doing the same in external scripts works reliable.

I just spend an hour searching for a problem. To avoid others to spend the hour again, here's a little "special behaviour" of shell execution in ELOG you should know about:

If you want to do redirect to a file in a shell execution, put a space before and after the redirecting. The following does not work:

You will not get an error message, but the file is not created. But if you add a space it will work as expected:

It is not really a bug; if you know about it, then it is not a big deal: hence this entry here. I saw this behavior on a Linux RHEL7 system.

In case you are wondering: I use this to create a default for the field CampaignID, to be used for new entries in combination with a Preset:

My config files are often large and use some shell scripting. In these shell scripting part I interact with files and execute scripts in specific folder.
It would be nice to have the ability to define for example a file path once and reuse it in different parts of the config file. It would help to make config files cleaner. For example:

Hi,

I am trying to access elog through a python client (https://github.com/paulscherrerinstitute/py_elog) and found a strage strange behavior which may be related server side problem. The python script generates get/post messages via the python requests library. This works fine so far and I can view and post messages. However, if a wrong user/password is provided, the server still returns HTTP status code "200 OK", although login failed. Instead, it should return something like "401 Unauthorized". This behavior later causes problems since the python client thinks login was successful. After experimenting around I think this could be caused by a server side misconfiguration. Any ideas?

I am not sure if this imformation is important: We use LDAP as user/password provider for elog.

Hi,

I am trying to access elog through a python client (https://github.com/paulscherrerinstitute/py_elog) and found a strage strange behavior which may be related server side problem. The python script generates get/post messages via the python requests library. This works fine so far and I can view and post messages. However, if a wrong user/password is provided, the server still returns HTTP status code "200 OK", although login failed. Instead, it should return something like "401 Unauthorized". This behavior later causes problems since the python client thinks login was successful. After experimenting around I think this could be caused by a server side misconfiguration. Any ideas?

I am not sure if this imformation is important: We use LDAP as user/password provider for elog.

Seems like I've discovered another bug here related to umlauts in my name. :D 

I was submitting this post and forgot to put an icon. Elog seems to have saved a copy of my message, which I could not edit since my username does not match the bugged name saved for this message.