public inbox for gnats-devel@sourceware.org
 help / color / mirror / Atom feed
* email follow up to GNATS PRs
@ 2003-03-03 15:10 ` Elena
  2003-03-10 23:10   ` Hans-Albert Schneider
  0 siblings, 1 reply; 3+ messages in thread
From: Elena @ 2003-03-03 15:10 UTC (permalink / raw)
  To: help-gnats

Hi,

Can anyone point me to the right direction....

We have gnats 4.0 beta version. When someone sends email follow up to 
gnats, following happens:

1. file "gnatsBlah" created in the gnats-queue directory
2. Short thereafter file disappears from the the gnats-queue directory (I 
think it is getting parsed correctly than)
3. Gnats sends email confirmation "The following reply was made to PR 
.....; it has been noted by GNATS." with a body of follow up.
4. Nothing else happens

I was expecting that at #4 I would see follow up appended to the PR's 
audit-trail, but this never happens. Any ideas of what may be wrong? I have 
checked file permissions, GNATS can write to PRs. Also, audit-trail entries 
working otherwise (State change, etc)

Any help is much appreciated!
-elena





_______________________________________________
Help-gnats mailing list
Help-gnats@gnu.org
http://mail.gnu.org/mailman/listinfo/help-gnats

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: email follow up to GNATS PRs
  2003-03-03 15:10 ` email follow up to GNATS PRs Elena
@ 2003-03-10 23:10   ` Hans-Albert Schneider
  2003-03-11  3:02     ` Elena
  0 siblings, 1 reply; 3+ messages in thread
From: Hans-Albert Schneider @ 2003-03-10 23:10 UTC (permalink / raw)
  To: Elena; +Cc: help-gnats

>>>>> "Elena" == Elena  <elena@brooklyn-realty.com> writes:

    Elena> Hi, Can anyone point me to the right direction....

    Elena> We have gnats 4.0 beta version. When someone sends
    Elena> email follow up to gnats, following happens:

    Elena> 1. file "gnatsBlah" created in the gnats-queue
    Elena> directory 2. Short thereafter file disappears from the
    Elena> the gnats-queue directory (I think it is getting
    Elena> parsed correctly than) 3. Gnats sends email
    Elena> confirmation "The following reply was made to PR
    Elena> .....; it has been noted by GNATS." with a body of
    Elena> follow up.  4. Nothing else happens

    Elena> I was expecting that at #4 I would see follow up
    Elena> appended to the PR's audit-trail, but this never
    Elena> happens. Any ideas of what may be wrong? I have
    Elena> checked file permissions, GNATS can write to
    Elena> PRs. Also, audit-trail entries working otherwise
    Elena> (State change, etc)

    Elena> Any help is much appreciated!  -elena


Try the following:

a) temporarily remove the cron job (so 2. will not happen by
   accident)

b) send a new follow-up e-mail

c) when this mail is in the queue, manually run the
	queue-pr -r -d <database-name>
   from the cron job.  Watch for error messages, exit code,
   core file etc.:

	cd <database-dir>/gnats-queue
	queue-pr -r -d <database-name>

   This hopefully gives you some error messages that point you
   into the right direction.

   Next, check the exit status ("echo $?" for bash, ksh and friends,
   "echo $status" for (t)csh).

   You may want to look for a "core dump":
	ls -l *core*

   If there is a core file, use the "file" command to find out
   which application produced it (probably queue-pr) and try to
   find the code location of the crash with a debugger, e.g.:

	gdb /usr/local/libexec/gnats/queue-pr core

   I don't know whether you are a programmer, and I am not good
   at describing debuggers.  The debugger has a command that
   lists the stack frames; for GNU gdb this is "where".


d) If this doesn't help you, repeat b) and c), but add "-D" to the
   queue-pr command.  This switches debug messages on; try to get
   a clue from them.


e) re-activate the cron job.


If you get some new information, but cannot solve the problem,
please come back *to the list* with this info (not to me, I will
be on vacation soon; sorry.)


Please also state which operating system and hardware platform
you use ("uname -a").

If this is a Linux system, please also state the vendor (RedHat,
SuSE, Debian, Mandrake, ...) and vendor version (e.g., "RedHat
7.2"); some systems have a file /etc/<vendorname>-release (e.g.,
/etc/SuSE-release) that should contain this information.

Did you compile GNATS yourself, or did you use a pre-compiled
package?


Hans-Albert

-- 
Hans-Albert Schneider
Munich, Germany
EMail: Hans-Albert@HA-Schneider.de


_______________________________________________
Help-gnats mailing list
Help-gnats@gnu.org
http://mail.gnu.org/mailman/listinfo/help-gnats

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: email follow up to GNATS PRs
  2003-03-10 23:10   ` Hans-Albert Schneider
@ 2003-03-11  3:02     ` Elena
  0 siblings, 0 replies; 3+ messages in thread
From: Elena @ 2003-03-11  3:02 UTC (permalink / raw)
  To: Hans-Albert; +Cc: help-gnats

Hans-Albert,

Thank you so much for detailed description. I must apologize - as I have found the root of the problem - but failed to report it back to the list.

At one point I was looking for a way to allow for a text to be added to the audit-trail manually (append) and I have followed an example - which suggest to add "Add-to-Audit-Trail" parameter in dbconfig - and also entailed setting "Audit-Trail" parameter to "read-only" 

When I removed "Add-to-Audit-Trail"  parameter (no longer needed) I forgot to remove "read-only" from the "Audit-Trail". This what have caused the behavior I described.

Not sure - whether it is a bug - or desired behavior - but I though it was a little odd. If it was truly read-only - I would of not expected to have "on-change" entries getting appended to audit-trail - and they did. Not sure what "read-only" really means for this parameter - whether it read-only for "user" or "internal system" etc... but ever since I have removed it - everything seem to be working just fine.

Thanks again for your help!

P.S. I have compiled Gnats - running Linux Red Hat 6.2 with 2.4.18 kernel

-elena


At 11:53 PM 3/10/2003 +0100, Hans-Albert Schneider wrote:
>>>>>> "Elena" == Elena  <elena@brooklyn-realty.com> writes:
>
>    Elena> Hi, Can anyone point me to the right direction....
>
>    Elena> We have gnats 4.0 beta version. When someone sends
>    Elena> email follow up to gnats, following happens:
>
>    Elena> 1. file "gnatsBlah" created in the gnats-queue
>    Elena> directory 2. Short thereafter file disappears from the
>    Elena> the gnats-queue directory (I think it is getting
>    Elena> parsed correctly than) 3. Gnats sends email
>    Elena> confirmation "The following reply was made to PR
>    Elena> .....; it has been noted by GNATS." with a body of
>    Elena> follow up.  4. Nothing else happens
>
>    Elena> I was expecting that at #4 I would see follow up
>    Elena> appended to the PR's audit-trail, but this never
>    Elena> happens. Any ideas of what may be wrong? I have
>    Elena> checked file permissions, GNATS can write to
>    Elena> PRs. Also, audit-trail entries working otherwise
>    Elena> (State change, etc)
>
>    Elena> Any help is much appreciated!  -elena
>
>
>Try the following:
>
>a) temporarily remove the cron job (so 2. will not happen by
>   accident)
>
>b) send a new follow-up e-mail
>
>c) when this mail is in the queue, manually run the
>        queue-pr -r -d <database-name>
>   from the cron job.  Watch for error messages, exit code,
>   core file etc.:
>
>        cd <database-dir>/gnats-queue
>        queue-pr -r -d <database-name>
>
>   This hopefully gives you some error messages that point you
>   into the right direction.
>
>   Next, check the exit status ("echo $?" for bash, ksh and friends,
>   "echo $status" for (t)csh).
>
>   You may want to look for a "core dump":
>        ls -l *core*
>
>   If there is a core file, use the "file" command to find out
>   which application produced it (probably queue-pr) and try to
>   find the code location of the crash with a debugger, e.g.:
>
>        gdb /usr/local/libexec/gnats/queue-pr core
>
>   I don't know whether you are a programmer, and I am not good
>   at describing debuggers.  The debugger has a command that
>   lists the stack frames; for GNU gdb this is "where".
>
>
>d) If this doesn't help you, repeat b) and c), but add "-D" to the
>   queue-pr command.  This switches debug messages on; try to get
>   a clue from them.
>
>
>e) re-activate the cron job.
>
>
>If you get some new information, but cannot solve the problem,
>please come back *to the list* with this info (not to me, I will
>be on vacation soon; sorry.)
>
>
>Please also state which operating system and hardware platform
>you use ("uname -a").
>
>If this is a Linux system, please also state the vendor (RedHat,
>SuSE, Debian, Mandrake, ...) and vendor version (e.g., "RedHat
>7.2"); some systems have a file /etc/<vendorname>-release (e.g.,
>/etc/SuSE-release) that should contain this information.
>
>Did you compile GNATS yourself, or did you use a pre-compiled
>package?
>
>
>Hans-Albert
>
>-- 
>Hans-Albert Schneider
>Munich, Germany
>EMail: Hans-Albert@HA-Schneider.de 



_______________________________________________
Help-gnats mailing list
Help-gnats@gnu.org
http://mail.gnu.org/mailman/listinfo/help-gnats

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2003-03-11  3:02 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
     [not found] <message <5.1.0.14.2.20030303093522.04240c30@mail.reefedge.com>
2003-03-03 15:10 ` email follow up to GNATS PRs Elena
2003-03-10 23:10   ` Hans-Albert Schneider
2003-03-11  3:02     ` Elena

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).