Re: Top Text and Bottom Text only show "text" --- no files, posted by Stefan Ritt on Fri Sep 22 08:21:39 2006
|
Steve Jones wrote: | Stefan, I found the source of the problem. When you moved some files to "logbook_dir" you also told the code to look in "logbook_dir" for top and bottom text files. The documentation indicates that the location dir should be "resource_dir". |
Yepp. I changed the documentation. Note that you can also specify an absolute path, like
/usr/local/elog/top.html
If the file name starts with a "/" (under Unix), the full path is taken instead of looking in the logbook directory. |
Re: Email substitution quit working, posted by Stefan Ritt on Fri Sep 22 08:50:08 2006
|
David Spindler wrote: |
David Spindler wrote: | I have recently upgraded from the 2.6.0-beta (I believe) to 2.6.2-1699. I just found out that on the day I upgraded, email substitution has stopped working. I have checked the discussion area and all the documentation and do not see any clues. The debug_log.txt file shows that the emails are being processed, but the fields are not being substituted correctly.
Thanks, in advance for any help,
David
|
I have not been able to find anything wrong in my config file, so I replaced 2.6.2 with 2.6.1 (apparently that was what I was running last, not the 2.6.0-beta) and my troubles have disappeared.
Anybody have any idea what has happened? |
This problem has been fixed in revision 1712. So I made an 2.6.2-2 for the Windows community just now. Please upgrade. |
Re: Attribute and checkbox, posted by Stefan Ritt on Fri Sep 22 08:55:22 2006
|
Well, what we use for such kind of check lists is the following:
Use Preset test = file.txt where file.txt contains exactly the ASCII code you mentioned, like
[ job done] [ customer ] [ activity ] [ What should be done ]
[] BLA Check job1 Backup mail check
[] BLA1 Check job2 check error report
[] Bla2 Check job3 check MQ jobs
So one can put the "X" in the "[]" manually. I know, the individual lines are no real elog attributes, but it works nicely for some check lists. Drawback might be that a use can edit the whole text, but sometimes I find that's even a plus, since the user might add some comments. |
Re: Shell execution generating error, posted by Steve Jones on Fri Sep 22 19:31:15 2006
|
Stefan Ritt wrote: |
Steve Jones wrote: | When started as root *but not running as a daemon* shell execution results in the following errors that are sent to Standard Error:
Cannot restore original GID/UID.
Cannot remove pidfile "/var/run/cr-elogd.pid"
; Permission denied
Cannot restore original GID/UID.
Cannot remove pidfile "/var/run/cr-elogd.pid"
; Permission denied
|
The "/var/run/elogd.pid" file is created from elogd to indicate under which PID it is running. If you run elogd once under root, this file then belongs to root. If you afterwards run it under a user account, it cannot delete or change the file belonging to root. In that case, just delete that file manually. |
Quote: |
When a process starts via the normal startup process it is started as root then the process changes to run as nobody -- so the pid file will always be owned by root. Yes? Then, shell commands wil not be able to deal with the pid file, right? Why would the shell exec want to deal with the PID file anyway?
Just curious. As long as this does not pose a problem then I will nto worry about it.
|
|
Re: SVN1714 will not run in 'daemon" mode on Solaris8, posted by Steve Jones on Fri Sep 22 19:32:45 2006
|
Stefan Ritt wrote: |
Steve Jones wrote: | On Solaris, SVN1714 will not go into daemon mode. Running the compiled version under TRUSS (which provides a dump of every system call) and shows precisely where elog is failing. I have attached two TRUSS outputs: one where it errors out and the other where it runs but "interactively". Both runs are as root, simply one with and one without the "-D" |
The "one where it errors out" does not look like an error. It does the "fork()" at the end and the main thread ends, that's how it's supposed to be. |
Quote: |
Ok, what this tells me is I need to get TRUSS to follow the fork - which I think I can do. The behavior, however, is that elog never goes into daemon mode after that fork.
More info to follow.
|
|
Re: "Supress Email Notification" checkbox, posted by Steve Jones on Fri Sep 22 19:42:21 2006
|
Stefan Ritt wrote: |
Steve Jones wrote: |
Steve Jones wrote: | Is there a way to allow email notification but have the "Suppress Email Notification" checked by default? We had complaints about the enormous amount of email so I turned off email notification for edits/replys, but there are times that I would like to allow email to be sent out -- this would imply allowing email by unchecking the "Suppress" box. |
What I think would be needed is:
Suppress Email on edit = 0|1|2
^
Where a '2' denotes that the "Suppress Email Notification" shows up but *checked* |
Just to a
Suppress default = 1 |
Quote: |
Not quite what I am looking for. What I need is:
"New" Logs
- Always send an email -- do not provide an option to check the box.
-- use Suppress Default = 2
"Edits and Replys"
- Provide the option to send an email, but by default the "suppress" box is checked
-- need Suppress Email on edit = 2
|
|
Re: SVN1714 will not run in 'daemon" mode on Solaris8, posted by Steve Jones on Fri Sep 22 22:12:18 2006
|
Stefan Ritt wrote: |
Steve Jones wrote: | On Solaris, SVN1714 will not go into daemon mode. Running the compiled version under TRUSS (which provides a dump of every system call) and shows precisely where elog is failing. I have attached two TRUSS outputs: one where it errors out and the other where it runs but "interactively". Both runs are as root, simply one with and one without the "-D" |
The "one where it errors out" does not look like an error. It does the "fork()" at the end and the main thread ends, that's how it's supposed to be. |
Quote: | Ok, I got it. I've attached the TRUSS output where we follow the fork. It appears that elogd cannot open any of the specified files then gives up. What was throwing me is no error output, even to STDERR. When I run the same but without the -D flag the files are opened fine.
There are also strange system calls that differ, and I thought it might be due to the setuid(60001) -nobody- but the the non-daemn mode also sets to nobody and works fine. |
|
Re: Invalid addresses block email notifications, posted by Stefan Ritt on Mon Sep 25 18:23:54 2006
|
Dimitrios Tsirigkas wrote: | Hi Stefan,
We have had problems with some users who subscribe for elog notifications using fake email addresses. When this happens, e-mail notifications stop working for all users and I have to find out which is the problematic email address and fix it myself. Is this a known problem? Is there a fix in a version newer than 2.6.1?
Thanks,
Dimitris |
Hi Dimitris,
in fact I have the same problem here with the forum. So far I did not find a clever solution, but I have some thought on how to minimize it:
- Do the self registration by sending a key to the registered email address. Only by using that key the account can be activated. This way at least on ensures a proper email address initially. There are however ways to temporaryly generate email accounts, which kind of bypasses this system. Furthermore, if an email address becomes invalid because the domain is switched off, this does not help.
- Email at least an error notification to the admin in case of failure.
You would say "why not removing the invalid email address"? Well that's kind of hard, since I send the email notification to all recipients in one email. Therefor I cannot figure out which is the offending email address. If I would send the emails separately to everyone, this can easily take minutes, which in turn would block the email server for that time. A solution would be a multithreaded server, which in fact is planned, but will not come so soon. If you come up with another clever solution please let me know. |
|