Skip to Content
Browse documentation

Reading Logs and Managing Retention

Read bounded Eenos log excerpts, correlate incidents, diagnose missing files and distinguish text-log rotation from database history cleanup.

Use an authorized administrator shell for server-wide diagnostics. Start with the incident time, timezone, account/domain, action and any message, request or job identifier.

Read a small time-relevant excerpt

tail -n 100 /var/log/eenos/eenos.log
tail -n 100 /var/log/eenos/eenosweb-wap.log

For new lines while reproducing a known issue:

tail -n 50 -F /var/log/eenos/eenostask.log

Stop following with Ctrl+C. -F follows by filename and retries across file replacement; use it only for a log you have identified. Output can include unrelated users' activity.

Use bounded journal queries for a relevant installed unit, for example:

journalctl -u eenos-mailtrack.service --since '1 hour ago' --no-pager -n 100

Journal retention and access are governed separately by the host. Different components use different timestamp formats; normalize their timezones before comparing events.

If a log is missing or empty

  1. Confirm the component is installed and that you selected the relevant panel, service and operation.
  2. Check its configured path. AppForge supports a logfile override, and Exim's source path is discovered from its configuration.
  3. Check the service journal for startup failures before file logging began.
  4. Check rotated files for the incident interval, including compressed archives where present.
  5. Check filesystem free space and access failures. Do not loosen permissions or change ownership just to make a viewer work.

A quiet or not-yet-used component may have no messages. A file being present or its final line saying that a command started does not prove successful completion.

Rotation is component-specific

Log familyRetention mechanism
Common Eenos operational logsShared file handlers; inspect the installed /etc/logrotate.d/eenos-* policy. No universal built-in size limit applies.
Panel Django logsWatched file handlers support external rotation.
MailTrack application logInternal bounded rotation and, where installed, a separate package logrotate rule.
AppForge main logWatched file handler and external logrotate policy.
AppForge job filesEligible job-history cleanup; independent of main-log rotation.
Website and service logsComponent/package configuration and scheduled rotation.

An installed rule is not proof that logrotate has run successfully. Review scheduling and errors before changing retention. Avoid applying an extra wildcard policy over files already managed by another rule.

Preserve evidence before cleanup

Activity Logs retention, MailTrack history cleanup and AppForge database maintenance affect their respective records. They do not provide one global switch for every server log.

Do not truncate active logs, delete SQLite sidecars or remove task directories to resolve a warning. Preserve the needed incident interval and use the supported component controls for deliberate cleanup. Manually deleting an open file may not free disk space until its writer closes it.

Prepare a useful support excerpt

Include the component, version, operation, incident time/timezone, affected identifier and relevant surrounding error lines. Redact passwords, tokens, keys and personal data while keeping correlation IDs consistent. Report whether the operation is still running or partially completed; do not attach entire server log folders or private task state.