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
When wazuh-remoted receives a keep-alive message from an agent, it processes it in two stages:
First, it receives the message, marks the agent as connected, updates the agent's status in the DB and queues the message to continue processing it later. This happens in function save_controlmsg.
Second, it unqueues the message, identifies the group to which the agent belongs, updates the DB and, if necessary, sends the new remote configuration. This happens in function read_controlmsg.
The problem with this is that, when a system is heavily loaded, it can take a long time to identify and save in the DB the group to which an agent belongs.
To avoid this, the logic that is in charge of guessing the group to which an agent belongs must be moved from stage two to stage one, in such a way that when an agent is in a connected state, it will always have some group assigned.
To do:
Update logic in Wazuh code.
Add UT for new functions.
Update documentation, if applies.
The text was updated successfully, but these errors were encountered:
Description
When
wazuh-remoted
receives a keep-alive message from an agent, it processes it in two stages:save_controlmsg
.read_controlmsg
.The problem with this is that, when a system is heavily loaded, it can take a long time to identify and save in the DB the group to which an agent belongs.
To avoid this, the logic that is in charge of guessing the group to which an agent belongs must be moved from stage two to stage one, in such a way that when an agent is in a
connected
state, it will always have some group assigned.To do:
The text was updated successfully, but these errors were encountered: