List in error_config but don't understand why

[mathieuroquefort]

Hello,

I have recently created a new list. It is the first creation since an upgrade from sympa 6.1.x to 6.2.48. We use families, and I create list through ssh requests : sympa.pl --add_list familly_name --input_file /etc/sympa/creation/new_list.xml

This list has been in error_config since its creation. But I just can't find why.
I've checked 3 times all the parameters in ‎⁨etc⁩/sympa⁩/families⁩/familly_name/param_constraint.conf and compared them with the actual parameters of the list, but they all fit.

The emails received to warn me indicate that I shall look at sympa logs for more information. But in the MySql table 'logs_table', there is nothing about this. All I find here about this list are my accesses to web interface, browsing all config pages to look for the problem.

How can I knom why this list has been set in error_config ?
I'm completely stuck...

[racke]

The 'logs_table' isn't the right place to look for the error messages. Sympa uses syslog: https://sympa-community.github.io/manual/install/configure-system-log.html. The actual location of the log file depends on your setup, but /var/log/sympa.log would be a good place to look for.

[mathieuroquefort]

Thank you.
That's better :)

But all I get there is many times this kind of thing :
sympa[29689]: err main::#849 > Sympa::Spindle::spin#95 > Sympa::Request::Handler::create_automatic_list::_twist#287 > Sympa::List::set_status_error_config#467 The list Sympa::List <[email protected]> is set in status error_config: no_respect_rules_family(diffusion_interne)
Without more details.

There is just one thing which seems strange : sympa[29689]: notice Sympa::Spindle::ProcessRequest::_twist() Processing Sympa::Request <action=create_automatic_list;[email protected];listname=collecteurs29>

I'm not sure what this create_automatic_list means. We use families, but not automatic families creation. There is no automatic_list_families in sympa.conf...

[ikedas]

Hi @mathieuroquefort ,

Could you please show us all of your family settings? See the description on family to find necessary settings.

[mathieuroquefort]

Hello @ikedas,
Thank you for interest.
Here they are.

config.tt2.txt
edit_list.conf.txt
param_constraint.conf.txt
new_diffint_collecteurs29.xml.txt
The last one is the instanciation file.

[ikedas]

@mathieuroquefort, I couldn't reproduce the problem using your configuration files.

Please show us all the log while sympa.pl was running, i.e. from when you hit the Enter key to when you noticed "error_config" status.

[mathieuroquefort]

Here it is.
The error_config status comes immediatly, there is nothing before. I've included all the logs about the list during 2 days.
collecteurs29-2019-12-20.log

[ikedas]

@mathieuroquefort,
Some log entries seem missing (possiblly due to syslog setting?).

Can you show the config file and config.0 file of collecteurs29 list? I want to compare them with mine.

[ikedas]

May be a duplicate of #841. Close for the present.

[mathieuroquefort]

Hello,
I'm opening this again because I'm facing the exact same problem.
My hoster solved the problem for collecteurs29 list (I don't know why) but we have the same one now with cfn list. I'd really like to understand what's happening here and not fix it list by list.

Here are all files which may be useful.
config.txt
config.0.txt
config.tt2.txt
edit_list.conf.txt
param_constraint.conf.txt

[mathieuroquefort]

It looks like commenting doesn't re-open the ticket.
Shall I create a new one ?

[ikedas]

@mathieuroquefort, please see #841. And

• if you think it the different issue, reopen this ticket (List in error_config but don't understand why #835) and post additional evidence here.
• if you think both the same issue, you need not reopen either.

[ikedas]

Hello,
I'm opening this again because I'm facing the exact same problem.
My hoster solved the problem for collecteurs29 list (I don't know why) but we have the same one now with cfn list. I'd really like to understand what's happening here and not fix it list by list.

Here are all files which may be useful.
config.txt
config.0.txt
config.tt2.txt
edit_list.conf.txt
param_constraint.conf.txt

config.tt2:

[% IF creation_email -%]
  email [% creation_email %]
[% END %]

It's broken. Resulting config file will cause error.

[mathieuroquefort]

Oh.
But what shall I do, then ?
I don't find any explanation about this on https://sympa-community.github.io/manual/customize/basics-families.html
Is it a deprecated parameter ?

[mathieuroquefort]

Also, this list has been instanciated long ago. config.tt2 is used only during instanciation I think. And many others lists have been instanciated with the same file and are not in error_config.

[ikedas]

config.tt2:

[% IF creation_email -%]
  email [% creation_email %]
[% END %]

Looking at templates bundled in Sympa distribution, it shoud be

creation
  date_epoch [% creation.date_epoch %]
  date       [% creation.date %]
[% IF creation_email -%]
  email      [% creation_email %]
[% END -%]

Recent version of Sympa complains to broken configuration. Earlier version looks ignoring some unreasonable things.

[mathieuroquefort]

Hello,

I've downloaded the last source, and in the file doc⁩/samples⁩/intranet⁩/create_list_templates/ ⁨intranet_list/config.tt2⁩ I see this :

[% IF creation_email -%]
  email [% creation_email %]
[% END %]

It is similar to the code that we currently use.
There is no[% END -%]

So, I don't understand why you consider my request as "invalid".

Is it the missing

creation
  date_epoch [% creation.date_epoch %]
  date       [% creation.date %]

?
In the config file of a recently created list, I have this :

creation
email [email protected]
date_epoch 1580143634

Exactly like for a very old list still worling.
So, it doesn't seem to be a problem.

[mathieuroquefort]

But maybe the place here to discuss this is not the good one and I should rather write to [email protected] ?

[mathieuroquefort]

I have finally solved the problem by re-instantiating the list.
I don't know if this %] instead of -%] was the actual cause. I doubt it, as it is effective only during instantiation.

Some more information explaining why a list is in error would be of great help.