Periodically checks of log

[Franz]

Periodically checks of log would be very useful. This means log-files are checked periodically (e.g. every 120 seconds) for any search pattern.

Example Use Case: When there is an Exception, leave a comment in a specific ticket; within this ticket users can add themselfs as "CC", so they will be automatically informed by email (if email is configured correctly).

Needed configuration in trac.ini:

[logwatcher]
# check interval in seconds, defaults to 120; 0 deactivates periodical checks
check_interval = 120
# amount of failures before entry will be deactivated
failure_limit = 10

Needed table:

create table logwatcher (
  id serial,
  name text,
  log_destination text,
  search_pattern text,
  is_regex bool,
  is_casesens bool,
  surround_lines int,
  last_check timestamp,
  last_line int,
  ticket bigint,
  active bool,
  failures int
);

Description of table logwatcher:

• ticket: if ticket ID is specified, leave comment at ticket; opt: if it is null, create new ticket
• failures: if it has failed to execute search (e.g. ticket ID is not valid, file not found, etc.), increase failures; when failures reach a specific amout (e.g. 10) deactivate this entry