Skip to content

Column filters

After a search finishes (or while results are available in a session), Column filters let you narrow the result table using the named capturing groups from your regex pattern.

These filters run on the already-downloaded session data (fast local filtering). They do not re-run the remote SSH search.

How columns appear

  1. The active directory/preset uses an extraction regex with capturing groups.
  2. Each named group becomes a column (e.g. timestamp, level, logger, thread, message).
  3. Rows are sorted chronologically by the timestamp group (oldest → newest). Lines with unparseable timestamps sort last.

Filterable vs timestamp

ColumnRole
timestampPrimary sort key. Not offered as a normal Combobox value filter in the column filter bar.
Other named groupsEach gets a filter control. Open it to pick suggested values or type a custom value.

Using a column filter

  1. Open the filter for a column (e.g. level).
  2. Optionally type in Search values to narrow suggestions (distinct values from the current session).
  3. Select one or more suggested values and/or enter a Custom value.
  4. Multiple selected values in the same column mean OR.
  5. Click Apply filters.
  6. Clear a column with Clear {column} filter when needed.

You can also toggle column visibility (Show column …) without clearing filters.

Relation to Complex search mode

  • Complex mode (on the search form) builds the remote pattern from group values before/during search.
  • Column filters refine the local result set after data is in the session.
  • Values from Complex mode may seed column filters; you can still change filters afterward without re-searching.

Examples

GoalFilter
Only errorslevelERROR
One consumer threadthread333 or [JMSConsumer-333] (substring-style match depending on stored cell text)
Errors or warningslevel → select ERROR and WARN (OR within the column)
Narrow logger packagelogger → custom value pl.npsesystem

If filters return nothing, use Clear on each active column or widen to a broader value. Also confirm the regex actually populated that column for matching lines.

Related