One log entry and your nc appears broken

ℹ️ Support
, , , , , ,
1

Nextcloud version: Nextcloud Hub 9 (30.0.1) (Linux 6.8.0-47-generic aarch64)
Operating system and version: Ubuntu 24.04.1 LTS
Webserver: Caddy Reverse Proxy running Docker container AIO
PHP version: 8.2.24
New installation on a Hetzner cloud server.

Safety & setup warnings shows broken website as soon as a single error (e.g. wrong credentials during login) is logged.

There doesn’t seem to be a button to empty or confirm the log file. Somewhat annoyed, I have now set the log level to 4, but this seems to be an unsatisfactory solution.

The log rotation only seems to empty the log when it has exceeded a certain amount. Manually resetting the log with cat /dev/null > /var/www/html/data/nextcloud.log seems to be an unnecessarily complicated solution (especially when working with Docker containers). Workaround would be a cron job to empty the log, but then I might miss some important information.

Is this the first time you’ve seen this error?

No. Permanently replicable condition.

Steps to replicate it:

Create any kind or error which gets logged, e.g. …

  • Log out.
  • Use the login form but enter wrong credentials.
  • Log in with correct credentials.
  • Go to General settings. After running a health check, your site shows a warning due to a new entry in the log file.

The output of your config.php file (stripped down to the essential lines regarding logging):

  'loglevel' => 4,
  'log_type' => 'file',
  'logfile' => '/var/www/html/data/nextcloud.log',
  'log_rotate_size' => 10485760,
  'log.condition' =>
  array (
    'apps' =>
    array (
      0 => 'admin_audit',
    ),
  ),

Desired result: Some way to confirm the log file or prune logs.
I would like to see a green tick under Safety & setup warnings. :white_check_mark:

(By the way, the documentation seems to have a wrong information regarding log_rotate_size, stating it would have a default rotate size of 100 MB. In fact it is only 10 MB (10485760) in a new NC installation.)

2

Sorry if this comes across as a bit harsh, but this has been discussed a zillion times in the forums, and yes, it would be nice if we could somehow confirm the log entries directly from the UI so that the green tick appears again.

What I’m wondering, though, is whether anyone who’s so incredibly bothered by this largely cosmetic issue has ever opened a feature request on GitHub with concrete suggestions for what such a feature should look like, or maybe even submitted a code contribution as a pull request…?

If not, why don’t you start with openng an issue on GitHub and then link to it here? In the meantime, you might be able to mitigate the ‘problem’ at OS level, at least partly, with standard UNIX/Linux tools like logrotate, or just by trying not to be triggered by the warning every time you see it :wink:

3

You are right in everything you say.
My apologies for that. :man_bowing:t2:

4

This topic was automatically closed 8 days after the last reply. New replies are no longer allowed.