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
Corrected the get_procs() documentation, which incorrectly stated that only
processes on the responding node of a Highly Available cluster are returned
(#529).
New get_processes() returns the processes Connect is running for content,
using the GET /v1/system/processes endpoint added in Connect 2026.11.0. It
can filter by content_guid, owner_guid, and hostname. On older Connect
versions, it falls back to the unversioned GET /metrics/procs endpoint and
returns the same columns (#531).
get_procs() and the Connect$procs() method are deprecated in favor of get_processes(). get_procs() now returns the server's pid instead of
the row index, parses start_time and heartbeat as datetimes, and its
documentation lists the columns it actually returns (#532).
Breaking changes
Improved performance of API response parsing and pagination. As a result,
getter functions no longer use a fixed list of columns. Data frames now
contain the fields that the server returns. If your Connect version does not
return a field, that column is not in the data frame. Some column types can
also be different across Connect versions. For example, IDs can be integers on
older versions (#483).