You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This change collects information about compressed chunks and stores
them in an LRU fashion. It captures events at compression, query
and DML oparations. The query/DML operations maintain totals and
information about the last operation.
The chunk information is maintained in shared memory per database.
One can access the data via calling the
`_timescaledb_functions.observ_chunks(..)` function or by
selecting from the
`timescaledb_information.observ_chunks` view
the former allows the user to push down filters, so it is faster
and less lock contention.
The amount of shared memory is controlled by the GUC:
`timescaledb.observ_max_chunks`
If the GUC is set to zero, the feature is disabled. The default
value is 1024 and the maximum is 32768. The amount must be a
power of two number. The `observ_max_chunks` value controls
how many chunk we store information about. The memory consumption
is 300 bytes per entry.
0 commit comments